The rise of AI-generated image tools has made localized forgeries increasingly realistic, posing challenges for visual content integrity. Although recent efforts have explored localized AIGC detection, existing datasets predominantly focus on object-level forgeries while overlooking broader scene edits in regions such as sky or ground. To address these limitations, we introduce BR-Gen, a large-scale dataset of 150,000 locally forged images with diverse scene-aware annotations, which are based on semantic calibration to ensure high-quality samples. BR-Gen is constructed through a fully automated ``Perception-Creation-Evaluation'' pipeline to ensure semantic coherence and visual realism. In addition, we further propose NFA-ViT, a Noise-guided Forgery Amplification Vision Transformer that enhances the detection of localized forgeries by amplifying subtle forgery-related features across the entire image. NFA-ViT mines heterogeneous regions in images, i.e., potential edited areas, by noise fingerprints. Subsequently, attention mechanism is introduced to compel the interaction between normal and abnormal features, thereby propagating the traces throughout the entire image, allowing subtle forgeries to influence a broader context and improving overall detection robustness. Extensive experiments demonstrate that BR-Gen constructs entirely new scenarios that are not covered by existing methods. Take a step further, NFA-ViT outperforms existing methods on BR-Gen and generalizes well across current benchmarks.

Recent advancements in computer vision (CV) and large language models (LLMs) have spurred significant interest in multi-modal large language models (MLLMs), which aim to integrate visual and textual modalities for enhanced understanding and generation tasks. While much of the existing research focuses on optimizing projectors and LLMs to improve MLLM performance, a critical question remains underexplored: Has the full potential of visual features in MLLMs been realized? To address this question, we identify two key limitations in current MLLM architectures and propose vMLLM, a vision-enhanced MLLM designed to fully leverage the capabilities of visual features. vMLLM introduces two novel components: the Multi-level Aggregation Module (MAM) and the Intra- and inter-modal Enhancement Module (IEM). The MAM aggregates multi-layer features from the vision encoder, capturing both high-level semantic information and low-level spatial details, thereby enriching the visual representation. The IEM enhances visual features through intra- and inter-modal interactions, effectively suppressing irrelevant information while amplifying task-relevant features, leading to more robust multimodal understanding. We conduct extensive experiments on multiple benchmarks, evaluating vMLLM across diverse settings, including different vision encoders, training dataset scales, and varying sizes of LLMs. Our results demonstrate that vMLLM consistently achieves significant performance improvements, validating its effectiveness in harnessing the potential of visual features. These findings highlight the importance of optimizing visual feature extraction and interaction mechanisms in MLLMs, paving the way for more advanced multimodal AI systems..

Story visualization has gained increasing attention in artificial intelligence. However, existing methods still struggle with maintaining a balance between character identity preservation and text-semantics alignment, largely due to a lack of detailed semantic modeling of the story scene. To tackle this challenge, we propose a novel knowledge graph, namely Character-Graph (CG), which represents various story-related knowledge, including the characters, their attributes and the relationship. We then introduce StoryWeaver, an image generator that achieves Customization via Character-Graph (C-CG), capable of consistent story visualization with rich text semantics. To further improve the multi-character generation performance, we incorporate knowledge-enhanced spatial guidance (KE-SG) into StoryWeaver to precisely inject character semantics into generation. To validate the effectiveness of our proposed method, extensive experiments are conducted using a new benchmark called TBC-Bench. The experiments confirm that our StoryWeaver excels not only in creating vivid visual story plots but also in accurately conveying character identities across various scenarios with considerable storage efficiency, e.g., achieving an average increase of +9.03% DINO-I and +13.44% CLIP-T. Furthermore, ablation experiments are conducted to verify the superiority of each proposed module.

3D Referring Expression Segmentation (3D-RES) aims to segment point cloud scenes based on a given expression. However, existing 3D-RES approaches face two major challenges: feature ambiguity and intent ambiguity. Feature ambiguity arises from information loss or distortion during point cloud acquisition due to limitations such as lighting and viewpoint. Intent ambiguity refers to the model's equal treatment of all queries during the decoding process, lacking top-down task-specific guidance. In this paper, we introduce an Image-enhanced Prompt Decoding Network (IPDN), which leverages multi-view images and task-driven information to enhance the model's reasoning capabilities. To address feature ambiguity, we propose the Multi-view Semantic Embedding (MSE) module, which injects multi-view 2D image information into the 3D scene and compensates for potential spatial information loss. To tackle intent ambiguity, we designed a Prompt-Aware Decoder (PAD) that guides the decoding process by deriving task-driven signals from the interaction between the expression and visual features. Comprehensive experiments demonstrate that IPDN outperforms the state-of-the-art by 1.9 and 4.2 points in mIoU metrics on the 3D-RES and 3D-GRES tasks, respectively.

Face manipulation techniques have achieved significant advances, presenting serious challenges to security and social trust. Recent works demonstrate that leveraging multimodal models can enhance the generalization and interpretability of face forgery detection. However, existing annotation approaches, whether through human labeling or direct Multimodal Large Language Model (MLLM) generation, often suffer from hallucination issues, leading to inaccurate text descriptions, especially for high-quality forgeries. To address this, we propose Face Forgery Text Generator (FFTG), a novel annotation pipeline that generates accurate text descriptions by leveraging forgery masks for initial region and type identification, followed by a comprehensive prompting strategy to guide MLLMs in reducing hallucination. We validate our approach through fine-tuning both CLIP with a three-branch training framework combining unimodal and multimodal objectives, and MLLMs with our structured annotations. Experimental results demonstrate that our method not only achieves more accurate annotations with higher region identification accuracy, but also leads to improvements in model performance across various forgery detection benchmarks. Our Codes are available in https://github.com/skJack/VLFFD.git.

Image restoration (IR), a key area in computer vision, has entered a new era with deep learning. Recent research has shifted toward Selective State Space Models (Mamba) to overcome CNNs’ limited receptive fields and Transformers’ computational inefficiency. However, due to Mamba’s inherent one-dimensional scanning limitations, recent approaches have introduced multi-directional scanning to bolster inter-sequence correlations. Despite these enhancements, these methods still struggle with managing local pixel correlations across various directions. Moreover, the recursive computation in Mamba’s SSM leads to reduced efficiency. To resolve these issues, we exploit the mathematical congruences between linear attention and SSM within Mamba to propose a novel model, ACL, which leverages news designs to Activate the Capability of Linear attention for IR. ACL integrates linear attention blocks instead of SSM within Mamba, serving as the core component of encoders/decoders, and aims to preserve a global perspective while boosting computational efficiency. Furthermore, we have designed a simple yet robust local enhancement module with multi-scale dilated convolutions to extract both coarse and fine features to improve local detail recovery. Experimental results confirm that our ACL model excels in classical IR tasks such as de-raining and de-blurring, while maintaining relatively low parameter counts and FLOPs1.

Despite a big leap forward in capability, multimodal large language models (MLLMs) tend to behave like a sloth in practical use, i.e., slow response and large latency. Recent efforts are devoted to building tiny MLLMs for better efficiency, but the plethora of visual tokens still used limit their actual speedup. In this paper, we propose a powerful and fast tiny MLLM called FlashSloth. Different from previous efforts, FlashSloth focuses on improving the descriptive power of visual tokens in the process of compressing their redundant semantics. In particular, FlashSloth introduces embedded visual compression designs to capture both visually salient and instruction-related image information, so as to achieving superior multimodal performance with fewer visual tokens. Extensive experiments are conducted to validate the proposed FlashSloth, and a bunch of tiny but strong MLLMs are also comprehensively compared, e.g., InternVL2, MiniCPM-V2 and Qwen2-VL. The experimental results show that compared with these advanced tiny MLLMs, our FlashSloth can greatly reduce the number of visual tokens, training memory and computation complexity while retaining high performance on various VL tasks. Our code is released at: https://github.com/codefanw/FlashSloth.

The rapid development of AI-generated content (AIGC) technology has led to the misuse of highly realistic AI-generated images (AIGI) in spreading misinformation, posing a threat to public information security. Although existing AIGI detection techniques are generally effective, they face two issues: 1) a lack of human-verifiable explanations, and 2) a lack of generalization in the latest generation technology. To address these issues, we introduce a large-scale and comprehensive dataset, Holmes-Set, which includes the Holmes-SFTSet, an instruction-tuning dataset with explanations on whether images are AI-generated, and the Holmes-DPOSet, a human-aligned preference dataset. Our work introduces an efficient data annotation method called the Multi-Expert Jury, enhancing data generation through structured MLLM explanations and quality control via cross-model evaluation, expert defect filtering, and human preference modification. In addition, we propose Holmes Pipeline, a meticulously designed three-stage training framework comprising visual expert pre-training, supervised fine-tuning, and direct preference optimization. Holmes Pipeline adapts multimodal large language models (MLLMs) for AIGI detection while generating human-verifiable and human-aligned explanations, ultimately yielding our model AIGI-Holmes. During the inference stage, we introduce a collaborative decoding strategy that integrates the model perception of the visual expert with the semantic reasoning of MLLMs, further enhancing the generalization capabilities. Extensive experiments on three benchmarks validate the effectiveness of our AIGI-Holmes.

Recently, mixture of experts (MoE) has become a popular paradigm for achieving the trade-off between modal capacity and efficiency of multimodal large language models (MLLMs). Different from previous efforts, we are dedicated to exploring the dynamic experts in existing MLLMs and showing that a standard MLLM can also be a mixture of experts. However, achieving this target is still notoriously challenging. The well-trained MLLMs are more accustomed to the fixed pathway and a drastic change in its inference manner also greatly impedes its performance. To address these issues, we propose a novel dynamic expert routing method for existing MLLMs, termed Routing Experts (RoE), which can achieve example-dependent optimal path routing without obvious structure tweaks. Meanwhile, a new structure sparsity regularization is also introduced to force the well-trained MLLMs to learn more short-cut pathways. In addition, we also address the alignment of the training and inference of MLLMs in terms of network routing. To validate RoE, we apply it to a set of existing MLLMs, including LLaVA-1.5, LLaVA-HR and VILA, and conduct extensive experiments on a bunch of VL benchmarks. The experiment results not only show the effectiveness of our RoE in improving MLLMs' efficiency, but also yield obvious advantages over MoE-LLaVA in both performance and speed, e.g., an average performance gain of 3.3% on 5 benchmarks while being 1.61 times faster. Our code is anonymously released at https://github.com/DoubtedSteam/RoE

In existing multimodal large language models (MLLMs), image resolution plays a significant role for granular visual recognition. However, directly increasing image resolution leads to expensive computational cost for MLLMs. In this paper, we reveal that a combination of low- and high-resolution visual features can efficiently mitigate this shortcoming. Based on this principle, we propose a novel and efficient method for MLLMs, termed Mixture-of-Resolution Adaptation (MRA). In particular, MRA adopts two visual pathways for images of different resolutions, where high-resolution visual information is embedded into the low-resolution pathway via the novel mixture-of-resolution adapters (MR-Adapters). This design also greatly reduces the input sequence length of MLLMs. To validate MRA, we apply it to a recent MLLM called LLaVA, and term the new model LLaVA-HR. We conduct extensive experiments on 17 vision-language (VL) tasks, which show that LLaVA-HR outperforms existing MLLMs on 15 VL tasks, e.g., +5.2\% on TextVQA. More importantly, both training and inference of LLaVA-HR remain efficient with MRA, e.g., 20 training hours and faster inference speed than LLaVA-NeXT. Source codes are released at: https://github.com/luogen1996/LLaVA-HR.

Personalized image generation is an important topic in text-to-image generation, with multi-ID personalization drawing widespread attention. Recent advancements in multi-ID image generation have led to zero-shot generation. However, existing multi-ID personalized generation methods cannot satisfy the real-world requirements of storytelling image generation where individuals actively interact through various poses and expressions. In this paper, we propose InterID, a plug-and-play framework designed for interaction-improved multi-ID personalization. We introduce a semantic prior as the spatial control signal in decoding texts to images, alleviating the challenge of generating complex interactions. The prior is obtained by arranging the individual concepts parsed from the input text with the aid of large language models and it is then fused with corresponding ID features to guide the flow of ID-enhanced features toward the target area. These ID-enhanced features, combined with the original text features, condition the generation process. Moreover, our approach is highly adaptable and can be seamlessly integrated with other custom models fine-tuned from the same base model, functioning as a versatile plugin. Extensive experiments validate the effectiveness and robustness of our approach, showing its superiority in multi-ID personalization. Our Code is released at https://github.com/sisibeauty/InterID.

Despite growing interest in hallucination in Multimodal Large Language Models, existing studies primarily focus on single-image settings, leaving hallucination in multi-image scenarios largely unexplored. To address this gap, we conduct the first systematic study of hallucinations in multi-image MLLMs and propose MIHBench, a benchmark specifically tailored for evaluating object-related hallucinations across multiple images. MIHBench comprises three core tasks: Multi-Image Object Existence Hallucination, Multi-Image Object Count Hallucination, and Object Identity Consistency Hallucination, targeting semantic understanding across object existence, quantity reasoning, and cross-view identity consistency. Through extensive evaluation, we identify key factors associated with the occurrence of multi-image hallucinations, including: a progressive relationship between the number of image inputs and the likelihood of hallucination occurrences; a strong correlation between single-image hallucination tendencies and those observed in multi-image contexts; and the influence of same-object image ratios and the positional placement of negative samples within image sequences on the occurrence of object identity consistency hallucination. To address these challenges, we propose a Dynamic Attention Balancing mechanism that adjusts inter-image attention distributions while preserving the overall visual attention proportion. Experiments across multiple state-of-the-art MLLMs demonstrate that our method effectively reduces hallucination occurrences and enhances semantic integration and reasoning stability in multi-image scenarios.

The reasoning segmentation task involves segmenting objects within an image by interpreting implicit user instructions. Despite significant advancements made by existing approaches, they remain constrained by low perceptual resolution, as visual encoders are typically pre-trained at lower resolutions. Furthermore, simply interpolating the positional embeddings of visual encoders to enhance perceptual resolution yields only marginal performance improvements while incurring substantial computational costs. To address this, we propose HRSeg, an efficient model with high-resolution fine-grained perception. It features two key innovations: High-Resolution Perception (HRP) and High-Resolution Enhancement (HRE). The HRP module processes high-resolution images through cropping, integrating local and global features for multi-granularity quality. The HRE module enhances mask features by integrating fine-grained information from high-resolution images, refining their alignment with text features for precise segmentation. Extensive ablation studies validate the effectiveness of our modules, while comprehensive experiments on multiple benchmark datasets demonstrate HRSeg's superior performance. Code will be available at https://github.com/WeihuangLin/HRSeg.

In this paper, we study the visual redundancy problem of multimodal large language models (MLLMs) from the perspective of attention behaviors. Via extensive empirical experiments, we observe and conclude three main inference stages of MLLMs: (i) Early fusion between tokens is first accomplished quickly. (ii) Intra-modality modeling then comes to play. (iii) Multimodal reasoning resumes and lasts until the end of inference. In particular, we reveal that visual tokens will stop contributing to reasoning when the text tokens receive enough image information. Based on this observation, we propose an effective method to improve the efficiency of MLLMs, termed dynamic visual-token exit (DyVTE), which is orthogonal but collaborative to previous token-wise visual compression methods. To validate the efficacy of DyVTE, we apply it to a set of MLLMs, including LLaVA, VILA, EAGLE and InternVL. The experimental results not only show the effectiveness of our DyVTE in improving MLLMs' efficiency, e.g., DyVTE reduces the computation overhead of LLaVA-1.5 by up to 45.7% without performance drop, but also reveal a general pattern across multiple MLLMs, well facilitating the in-depth analysis of MLLMs. Our code is anonymously released at https://anonymous.4open.science/r/AnonymousDyVTE-26AB/.

The rising importance of 3D representation learning, pivotal in computer vision, autonomous driving, and robotics, is evident. However, a prevailing trend, which straightforwardly resorted to transferring 2D alignment strategies to the 3D domain, encounters three distinct challenges: (1) Information Degradation: This arises from the alignment of 3D data with mere single-view 2D images and generic texts, neglecting the need for multi-view images and detailed subcategory texts. (2) Insufficient Synergy: These strategies align 3D representations to image and text features individually, hampering the overall optimization for 3D models. (3) Underutilization: The fine-grained information inherent in the learned representations is often not fully exploited, indicating a potential loss in detail. To address these issues, we introduce JM3D, a comprehensive approach integrating point cloud, text, and image. Key contributions include the Structured Multimodal Organizer (SMO), enriching vision-language representation with multiple views and hierarchical text, and the Joint Multi-modal Alignment (JMA), combining language understanding with visual representation. Our advanced model, JM3D-LLM, marries 3D representation with large language models via efficient fine-tuning. Evaluations on ModelNet40 and ScanObjectNN establish JM3D's superiority. The superior performance of JM3D-LLM further underscores the effectiveness of our representation transfer approach.

Pre-training and fine-tuning have been the de-facto paradigm in vision-language domains. Along with the rapid growth of model sizes, fully fine-tuning these large-scale vision-language pre-training (VLP) models requires prohibitively expensive storage costs. To address this issue, recent advances in NLP offer a promising and efficient adaptation approach called LoRA, which aims to approximate the fine-tuning of large pre-trained model by updating low-rank parameters. Despite its effectiveness, we identify that LoRA suffers a large approximation error on VLP models and its optimization is also inefficient, which greatly limits its performance upper bound. In this paper, we mathematically prove that the approximation error of low-rank adaptation can be optimized by a new optimization objective, i.e., the weight distance between LoRA and fine-tuning. Based on this finding, we propose a novel PETL method for VLP models, namely momentum imitation learning (MoIL). Specifically, MoIL formulates PETL as a weight imitation learning process and directly optimize the approximation error bound of the low-rank adaptation. Based on this training scheme, we also explore a new hybrid approximation function to reduce the learning difficulty of low-rank adaptations. With these two novel designs, MoIL can greatly improve the optimization efficiency of the low-rank parameters on VLP models. We validate MoIL on three VLP models ranging from end-to-end network to two-stage network, and conduct extensive experiments on four VL tasks. Experimental results demonstrate superior performance and optimization efficiency of MoIL than existing PETL methods. For instance, by updating only 6.23% parameters, MoIL can even outperform full tuning by +2.3% on image-text matching task. Meanwhile, its inference efficiency and generalization ability is also validated by multiple VLP models, e.g., VLMO and VinVL.

Deep neural networks often suffer from poor generalization due to complex and non-convex loss landscapes. Sharpness-Aware Minimization (SAM) is a popular solution that smooths the loss landscape by minimizing the maximized change of training loss when adding a perturbation to the weight. However, indiscriminate perturbation of SAM on all parameters is suboptimal and results in excessive computation, double the overhead of common optimizers like Stochastic Gradient Descent (SGD). In this paper, we propose Sparse SAM (SSAM), an efficient and effective training scheme that achieves sparse perturbation by a binary mask. To obtain the sparse mask, we provide two solutions based on Fisher information and dynamic sparse training, respectively. We investigate the impact of different masks, including unstructured, structured, and $N$N:$M$M structured patterns, as well as explicit and implicit forms of implementing sparse perturbation. We theoretically prove that SSAM can converge at the same rate as SAM, i.e., $O(\log T/\sqrt{T})$O(logT/T) . Sparse SAM has the potential to accelerate training and smooth the loss landscape effectively. Extensive experimental results on CIFAR and ImageNet-1K confirm that our method is superior to SAM in terms of efficiency, and the performance is preserved or even improved with a perturbation of merely 50% sparsity.

Camouflaged Object Detection (COD) poses a significant challenge in computer vision, playing a critical role in applications. Existing COD methods often exhibit challenges in accurately predicting nuanced boundaries with high-confidence predictions. In this work, we introduce CamoDiffusion, a new learning method that employs a conditional diffusion model to generate masks that progressively refine the boundaries of camouflaged objects. In particular, we first design an adaptive transformer conditional network, specifically designed for integration into a Denoising Network, which facilitates iterative refinement of the saliency masks. Second, based on the classical diffusion model training, we investigate a variance noise schedule and a structure corruption strategy, which aim to enhance the accuracy of our denoising model by effectively handling uncertain input. Third, we introduce a Consensus Time Ensemble technique, which integrates intermediate predictions using a sampling mechanism, thus reducing overconfidence and incorrect predictions. Finally, we conduct extensive experiments on three benchmark datasets that show that: 1) the efficacy and universality of our method is demonstrated in both camouflaged and salient object detection tasks. 2) compared to existing state-of-the-art methods, CamoDiffusion demonstrates superior performance 3) CamoDiffusion offers flexible enhancements, such as an accelerated version based on the VQ-VAE model and a skip approach.

Panoptic Narrative Detection (PND) and Segmentation (PNS) are two challenging tasks that involve identifying and locating multiple targets in an image according to a long narrative description. In this paper, we propose a unified and effective framework called NICE that can jointly learn these two panoptic narrative recognition tasks. Existing visual grounding tasks use a two-branch paradigm, but applying this directly to PND and PNS can result in prediction conflict due to their intrinsic many-to-many alignment property. To address this, we introduce two cascading modules based on the barycenter of the mask, which are Coordinate Guided Aggregation (CGA) and Barycenter Driven Localization (BDL), responsible for segmentation and detection, respectively. By linking PNS and PND in series with the barycenter of segmentation as the anchor, our approach naturally aligns the two tasks and allows them to complement each other for improved performance. Specifically, CGA provides the barycenter as a reference for detection, reducing BDL's reliance on a large number of candidate boxes. BDL leverages its excellent properties to distinguish different instances, which improves the performance of CGA for segmentation. Extensive experiments demonstrate that NICE surpasses all existing methods by a large margin, achieving 4.1% for PND and 2.9% for PNS over the state-of-the-art. These results validate the effectiveness of our proposed collaborative learning strategy.

Despite the great success, most models in image captioning (IC) are still stuck in the dilemma of generating simple and non-discriminative captions. In this paper, we study this problem from the perspective of data augmentation and propose a novel method called Multi-modal Mixup ( M 3 ixup). Compared with the original Mixup strategy designed for image classification , the proposed M 3 ixup has three novel designs to mix IC samples from the aspects of visual features, sentence embeddings and loss values, respectively. In practice, M 3 ixup can not only enrich the diversity of IC training data , but also enforce the model to focus more on visual information for captioning, thereby alleviating the negative effect of dataset bias and addressing the issue of simple captioning. To validate M 3 ixup, we apply it to three baseline models and conduct extensive experiments on MS COCO. The experimental results demonstrate that our proposed M 3 ixup can not only improve the discriminability and quality of generated captions, but also help the baseline models obtain obvious performance gains, i.e. , improving the CIDEr scores of the state-of-the-art model from 133.8 to 135.3 on off-line testing and 135.4 to 137.1 on online testing.

The focus of Face Anti-Spoofing (FAS) is shifting toward improving generalization performance in unseen scenarios. Traditional methods employing adversarial learning and meta-learning aim to extract or decouple generalizable features to address these challenges. However, enhancing performance solely through facial features remains challenging without additional informative inputs. To address this, Vision-Language Models (VLMs) with robust generalization capabilities have recently been introduced to FAS. Despite their potential, these VLMs typically adopt a late alignment strategy, relying only on encoder output features for modality alignment, which largely neglects mutual guidance between modalities. To bridge this gap, inspired by recent advancements in prompt learning, we employ learnable prompts and masking as intermediaries to enhance interaction between text and visual modalities, enabling the extraction of more generalizable features. Specifically, we propose ME-FAS, a Modality-Enhanced cross-domain FAS model integrating Prompt Fusion Transfer (PFT) and Text-guided Image Masking (TIM). PFT facilitates the integration of text features with visual information, improving domain adaptability in alignment with the textual context. Meanwhile, TIM leverages text features to mask image patches, directing visual features toward critical generalizable facial information, such as the eyes and mouth. Comprehensive evaluations across multiple benchmarks and various visualizations demonstrate significant performance gains, validating the effectiveness of our proposed approach. Our code and models are available at https://github.com/clpbc/ME-FAS.

This article explores a novel dynamic network for vision and language (V&L) tasks, where the inferring structure is customized on the fly for different inputs. Most previous state-of-the-art (SOTA) approaches are static and handcrafted networks, which not only heavily rely on expert knowledge but also ignore the semantic diversity of input samples, therefore resulting in suboptimal performance. To address these issues, we propose a novel Dynamic Transformer Network (DTNet) for image captioning, which dynamically assigns customized paths to different samples, leading to discriminative yet accurate captions. Specifically, to build a rich routing space and improve routing efficiency, we introduce five types of basic cells and group them into two separate routing spaces according to their operating domains, i.e., spatial and channel. Then, we design a Spatial-Channel Joint Router (SCJR), which endows the model with the capability of path customization based on both spatial and channel information of the input sample. To validate the effectiveness of our proposed DTNet, we conduct extensive experiments on the MS-COCO dataset and achieve new SOTA performance on both the Karpathy split and the online test server. The source code is publicly available at https://github.com/xmu-xiaoma666/DTNet.

In recent times, automatic text-to-3D content creation has made significant progress, driven by the development of pretrained 2D diffusion models. Existing text-to-3D methods typically optimize the 3D representation to ensure that the rendered image aligns well with the given text, as evaluated by the pretrained 2D diffusion model. Nevertheless, a substantial domain gap exists between 2D images and 3D assets, primarily attributed to variations in camera-related attributes and the exclusive presence of foreground objects. Consequently, employing 2D diffusion models directly for optimizing 3D representations may lead to suboptimal outcomes. To address this issue, we present X-Dreamer, a novel approach for high-quality text-to-3D content creation that effectively bridges the gap between text-to-2D and text-to-3D synthesis. The key components of X-Dreamer are two innovative designs: Camera-Guided Low-Rank Adaptation (CG-LoRA) and Attention-Mask Alignment (AMA) Loss. CG-LoRA dynamically incorporates camera information into the pretrained diffusion models by employing camera-dependent generation for trainable parameters. This integration makes the 2D diffusion model camera-sensitive. AMA loss guides the attention map of the pretrained diffusion model using the binary mask of the 3D object, prioritizing the creation of the foreground object. This module ensures that the model focuses on generating accurate and detailed foreground objects. Extensive evaluations demonstrate the effectiveness of our proposed method compared to existing text-to-3D approaches. Our project webpage: https://anonymous-11111.github.io/ . Our code is available at https://github.com/xmu-xiaoma666/X-Dreamer .

Recent advancements in single-stage Panoptic Narrative Grounding (PNG) have demonstrated significant potential. These methods predict pixel-level masks by directly matching pixels and phrases. However, they often neglect the modeling of semantic and visual relationships between phrase-level instances, limiting their ability for complex multi-modal reasoning in PNG. To tackle this issue, we propose XPNG, a “differentiation-refinement-localization” reasoning paradigm for accurately locating instances or regions. In XPNG, we introduce a Semantic Context Convolution (SCC) module to leverage semantic priors for generating distinctive features. This well-crafted module employs a combination of dynamic channel-wise convolution and pixel-wise convolution to embed semantic information and establish inter-object relationships guided by semantics. Subsequently, we propose a Visual Context Verification (VCV) module to provide visual cues, eliminating potential space biases introduced by semantics and further refining the visual features generated by the previous module. Extensive experiments on PNG benchmark datasets reveal that our approach achieves state-of-the-art performance, significantly outperforming existing methods by a considerable margin and yielding a 3.9-point improvement in overall metrics. Our codes and results are available at our project webpage: https://github.com/TianyuGoGO/XPNG.

Referring 3D instance segmentation is a challenging task aimed at accurately segmenting a target instance within a 3D scene based on a given referring expression. However, previous methods have overlooked the distinct roles played by different words in referring expressions. Additionally, they have failed to incorporate the positional relationship within referring expressions with the spatial correlations in 3D scenes. To alleviate these issues, we present a novel model called X-RefSeg3D, which constructs a cross-modal graph for the input 3D scene and unites textual and spatial relationships for reasoning via graph neural networks. Our approach begins by capturing object-specific text features, which are then fused with the instance features to construct a comprehensive cross-modal scene graph. Subsequently, we integrate the obtained cross-modal features into graph neural networks, leveraging the K-nearest algorithm to derive explicit instructions from expressions and factual relationships in scenes. This enables the effective capture of higher-order relationships among instances, thereby enhancing feature fusion and facilitating reasoning. Finally, the refined feature undergoes a matching module to compute the ultimate matching score. Experimental results on ScanRefer demonstrate the effectiveness of our method, surpassing previous approaches by a substantial margin of +3.67% in terms of mIOU.

This work is oriented toward the task of open-set Human Object Interaction (HOI) detection. The challenge lies in identifying completely new, out-of-domain relationships, as opposed to in-domain ones which have seen improvements in zero-shot HOI detection. To address this challenge, we introduce a simple Disentangled HOI Detection (DHD) model for detecting novel relationships by integrating an open-set object detector with a Visual Language Model (VLM). We utilize a disentangled image-text contrastive learning metric for training and connect the bottom-up visual features to text embeddings through lightweight unary and pair-wise adapters. Our model can benefit from the open-set object detector and the VLM to detect novel action categories and combine actions with novel object categories. We further present the VG-HOI dataset, a comprehensive benchmark with over 17k HOI relationships for open-set scenarios. Experimental results show that our model can detect unknown action classes and combine unknown object classes. Furthermore, it can generalize to over 17k HOI classes while being trained on just 600 HOI classes.

In 3D Referring Expression Segmentation (3D-RES), the earlier approach adopts a two-stage paradigm, extracting segmentation proposals and then matching them with referring expressions. However, this conventional paradigm encounters significant challenges, most notably in terms of the generation of lackluster initial proposals and a pronounced deceleration in inference speed. Recognizing these limitations, we introduce an innovative end-to-end Superpoint-Text Matching Network (3D-STMN) that is enriched by dependency-driven insights. One of the keystones of our model is the Superpoint-Text Matching (STM) mechanism. Unlike traditional methods that navigate through instance proposals, STM directly correlates linguistic indications with their respective superpoints, clusters of semantically related points. This architectural decision empowers our model to efficiently harness cross-modal semantic relationships, primarily leveraging densely annotated superpoint-text pairs, as opposed to the more sparse instance-text pairs. In pursuit of enhancing the role of text in guiding the segmentation process, we further incorporate the Dependency-Driven Interaction (DDI) module to deepen the network's semantic comprehension of referring expressions. Using the dependency trees as a beacon, this module discerns the intricate relationships between primary terms and their associated descriptors in expressions, thereby elevating both the localization and segmentation capacities. Comprehensive experiments on the ScanRefer benchmark reveal that our model not only sets new performance standards, registering an mIoU gain of 11.7 points but also achieves a staggering enhancement in inference speed, surpassing traditional methods by 95.7 times. The code and models are available at https://github.com/sosppxo/3D-STMN.

Diffusion-based Image Editing (DIE) is an emerging research hot-spot, which often applies a semantic mask to control the target area for diffusion-based editing. However, most existing solutions obtain these masks via manual operations or off-line processing, greatly reducing their efficiency. In this paper, we propose a novel and efficient image editing method for Text-to-Image (T2I) diffusion models, termed Instant Diffusion Editing (InstDiffEdit). In particular, InstDiffEdit aims to employ the cross-modal attention ability of existing diffusion models to achieve instant mask guidance during the diffusion steps. To reduce the noise of attention maps and realize the full automatics, we equip InstDiffEdit with a training-free refinement scheme to adaptively aggregate the attention distributions for the automatic yet accurate mask generation. Meanwhile, to supplement the existing evaluations of DIE, we propose a new benchmark called Editing-Mask to examine the mask accuracy and local editing ability of existing methods. To validate InstDiffEdit, we also conduct extensive experiments on ImageNet and Imagen, and compare it with a bunch of the SOTA methods. The experimental results show that InstDiffEdit not only outperforms the SOTA methods in both image quality and editing results, but also has a much faster inference speed, i.e., +5 to +6 times. Our code available at https://anonymous.4open.science/r/InstDiffEdit-C306

Referring Remote Sensing Image Segmentation (RRSIS) is a new challenge that combines computer vision and natu-ral language processing. Traditional Referring Image Seg-mentation (RIS) approaches have been impeded by the com-plex spatial scales and orientations found in aerial imagery, leading to suboptimal segmentation results. To address these challenges, we introduce the Rotated Multi-Scale In-teraction Network (RMSIN), an innovative approach de-signed for the unique demands of RRSIS. RMSIN incorpo-rates an Intra-scale Interaction Module (IIM) to effectively address the fine-grained detail required at multiple scales and a Cross-scale Interaction Module (CIM) for integrating these details coherently across the network. Furthermore, RMSIN employs an Adaptive Rotated Convolution (ARC) to account for the diverse orientations of objects, a novel contribution that significantly enhances segmentation accu-racy. To assess the efficacy of RMSIN, we have curated an expansive dataset comprising 17,402 image-caption-mask triplets, which is unparalleled in terms of scale and vari-ety. This dataset not only presents the model with a wide range of spatial and rotational scenarios but also estab-lishes a stringent benchmark for the RRSIS task, ensuring a rigorous evaluation of performance. Experimental eval-uations demonstrate the exceptional performance of RM-SIN, surpassing existing state-of-the-art models by a signif-icant margin. Datasets and code are available at https://github.com/Lsan2401/RMSIN.

Referring Expression Segmentation (RES) is a challenging task in computer vision that involves segmenting image instances using textual descriptions. Conventional approaches suffer from the high cost of acquiring segmentation labels. To overcome this, we propose a novel learning task called Omni-supervised Referring Expression Segmentation (Omni-RES) which leverages unlabeled, fully labeled, and weakly labeled data, such as referring points or bounding boxes, for efficient RES training. Our approach is based on a teacher-student learning framework, where weak labels guide the selection and refinement of high-quality pseudo-masks for training, rather than serving as direct supervision signals. We tested Omni-RES on various state-of-the-art RES models and datasets, demonstrating its superiority over fully-supervised and semi-supervised methods. Remarkably, with just 10% fully labeled data, Omni-RES can match the performance of 100% supervised training. Additionally, it enables using large-scale vision-language datasets like Visual Genome for cost-effective RES training, setting a new state-of-the-art performance in RES, such as 80.66 on RefCOCO. Our code is released at: https://github.com/nineblu/omni-res

Text-to-3D-aware face (T3D Face) generation and manipulation is an emerging research hot spot in machine learning, which still suffers from low efficiency and poor quality. In this paper, we propose an ***E**nd-to-End **E**fficient and **E**ffective* network for fast and accurate T3D face generation and manipulation, termed $E^3$-FaceNet. Different from existing complex generation paradigms, $E^3$-FaceNet resorts to a direct mapping from text instructions to 3D-aware visual space. We introduce a novel *Style Code Enhancer* to enhance cross-modal semantic alignment, alongside an innovative *Geometric Regularization* objective to maintain consistency across multi-view generations. Extensive experiments on three benchmark datasets demonstrate that $E^3$-FaceNet can not only achieve picture-like 3D face generation and manipulation, but also improve inference speed by orders of magnitudes. For instance, compared with Latent3D, $E^3$-FaceNet speeds up the five-view generations by almost 470 times, while still exceeding in generation quality. Our code is released at <https://github.com/Aria-Zhangjl/E3-FaceNet>.

Recent advancements in automatic 3D avatar generation guided by text have made significant progress. However, existing methods have limitations such as oversaturation and low-quality output. To address these challenges, we propose X-Oscar, a progressive framework for generating high-quality animatable avatars from text prompts. It follows a sequential "Geometry→Texture→Animation" paradigm, simplifying optimization through step-by-step generation. To tackle oversaturation, we introduce Adaptive Variational Parameter (AVP), representing avatars as an adaptive distribution during training. Additionally, we present Avatar-aware Score Distillation Sampling (ASDS), a novel technique that incorporates avatar-aware noise into rendered images for improved generation quality during optimization. Extensive evaluations confirm the superiority of X-Oscar over existing text-to-3D and text-to-avatar approaches. Our anonymous project page: https://anonymous1440.github.io/.

The issue of hallucinations is a prevalent concern in existing Large Vision-Language Models (LVLMs). Previous efforts have primarily focused on investigating object hallucinations, which can be easily alleviated by introducing object detectors. However, these efforts neglect hallucinations in inter-object relationships, which is essential for visual comprehension. In this work, we introduce R-Bench, a novel benchmark for evaluating Vision Relationship Hallucination. R-Bench features image-level questions that focus on the existence of relationships and instance-level questions that assess local visual comprehension. We identify three types of relationship co-occurrences that lead to hallucinations: relationship-relationship, subject-relationship, and relationship-object. The visual instruction tuning dataset's long-tail distribution significantly impacts LVLMs' understanding of visual relationships. Additionally, our analysis reveals that current LVLMs tend to overlook visual content, overly rely on the common sense knowledge of Large Language Models (LLMs), and struggle with spatial relationship reasoning based on contextual information.

In this paper, we introduce SemiRES, a semi-supervised framework that effectively leverages a combination of labeled and unlabeled data to perform RES. A significant hurdle in applying semi-supervised techniques to RES is the prevalence of noisy pseudo-labels, particularly at the boundaries of objects. SemiRES incorporates the Segment Anything Model (SAM), renowned for its precise boundary demarcation, to improve the accuracy of these pseudo-labels. Within SemiRES, we offer two alternative matching strategies: IoU-based Optimal Matching (IOM) and Composite Parts Integration (CPI). These strategies are designed to extract the most accurate masks from SAM’s output, thus guiding the training of the student model with enhanced precision. In instances where a precise mask cannot be matched from the available candidates, we develop the Pixel-Wise Adjustment (PWA) strategy, guiding the student model’s training directly by the pseudo-labels. Extensive experiments on three RES benchmarks—RefCOCO, RefCOCO+, and G-Ref reveal its superior performance compared to fully supervised methods, especially in low-data scenarios. Remarkably, with only 1% labeled data, our SemiRES outperforms the supervised baseline by a large margin, e.g. +18.64% gains on RefCOCO val set.

Visual grounding is a task of locating the object referred by a natural language description. To reduce annotation costs, recent researchers are devoted into one-stage weakly supervised methods for visual grounding, which typically adopt the anchor-text matching paradigm. Despite the efficiency, we identify that anchor representations are often noisy and insufficient to describe object information, which inevitably hinders the vision-language alignments. In this paper, we propose a novel query-based one-stage framework for weakly supervised visual grounding, namely QueryMatch. Different from previous work, QueryMatch represents candidate objects with a set of query features, which inherently establish accurate one-to-one associations with visual objects. In this case, QueryMatch re-formulates weakly supervised visual grounding as a query-text matching problem, which can be optimized via the query-based contrastive learning. Based on QueryMatch, we further propose an innovative strategy for effective weakly supervised learning, namely Active Query Selection (AQS). In particular, AQS aims to enhance the effectiveness of query-based contrastive learning by actively selecting high-quality query features. Through this strategy, AQS can greatly benefit the weakly supervised learning of QueryMatch. To validate our approach, we conduct extensive experiments on three benchmark datasets of two grounding tasks, i.e., referring expression comprehension (REC) and segmentation (RES). Experimental results not only show the state-of-art performance of QueryMatch in two tasks, e.g., over +5% [email protected] on RefCOCO in REC and over +20% mIOU on RefCOCO in RES, but also confirm the effectiveness of AQS in weakly supervised learning. Source codes are available at https://github.com/TensorThinker/QueryMatch.

Recently, Segment Anything Model (SAM) has become a research hotspot in the fields of multimedia and computer vision, which exhibits powerful yet versatile capabilities on various (un) conditional image segmentation tasks. Although SAM can support different types of segmentation prompts, we note that, compared to point- and box-guided segmentations, it performs much worse on text-instructed tasks, e.g., referring image segmentation (RIS). In this paper, we argue that deep text instruction tuning is key to mitigate such shortcoming caused by the shallow fusion scheme in its default light-weight mask decoder. To address this issue, we propose two simple yet effective deep instruction tuning (DIT) methods for SAM, one is end-to-end and the other is layer-wise. With minimal modifications, DITs can directly transform the image encoder of SAM as a stand-alone vision-language learner in contrast to building another deep fusion branch, maximizing the benefit of its superior segmentation capability. Extensive experiments on three highly competitive benchmark datasets of RIS show that a simple end-to-end DIT can improve SAM by a large margin, while the layer-wise DIT can further boost the performance to state-of-the-art with much less data and training expenditures. Our code is released at: https://github.com/wysnzzzz/DIT.

3D Referring Expression Segmentation (3D-RES) is dedicated to segmenting a specific instance within a 3D space based on a natural language description. However, current approaches are limited to segmenting a single target, restricting the versatility of the task. To overcome this limitation, we introduce Generalized 3D Referring Expression Segmentation (3D-GRES), which extends the capability to segment any number of instances based on natural language instructions. In addressing this broader task, we propose the Multi-Query Decoupled Interaction Network (MDIN), designed to break down multi-object segmentation tasks into simpler, individual segmentations. MDIN comprises two fundamental components: Text-driven Sparse Queries (TSQ) and Multi-object Decoupling Optimization (MDO). TSQ generates sparse point cloud features distributed over key targets as the initialization for queries. Meanwhile, MDO is tasked with assigning each target in multi-object scenarios to different queries while maintaining their semantic consistency. To adapt to this new task, we build a new dataset, namely Multi3DRes. Our comprehensive evaluations on this dataset demonstrate substantial enhancements over existing models, thus charting a new path for intricate multi-object 3D scene comprehension. The benchmark and code are available at https://github.com/sosppxo/MDIN.

The rapid progress in generative models has given rise to the critical task of AI-Generated Content Stealth (AIGC-S), which aims to create AI-generated images that can evade both forensic detectors and human inspection. This task is crucial for understanding the vulnerabilities of existing detection methods and developing more robust techniques. However, current adversarial attacks often introduce visible noise, have poor transferability, and fail to address spectral differences between AI-generated and genuine images. To address this, we propose StealthDiffusion, a framework based on stable diffusion that modifies AI-generated images into high-quality, imperceptible adversarial examples capable of evading state-of-the-art forensic detectors. StealthDiffusion comprises two main components: Latent Adversarial Optimization, which generates adversarial perturbations in the latent space of stable diffusion, and Control-VAE, a module that reduces spectral differences between the generated adversarial images and genuine images without affecting the original diffusion model's generation process. Extensive experiments show that StealthDiffusion is effective in both white-box and black-box settings, transforming AI-generated images into high-quality adversarial forgeries with frequency spectra similar to genuine images. These forgeries are classified as genuine by advanced forensic classifiers and are difficult for humans to distinguish.

The rapid progress of Deepfake technology has made face swapping highly realistic, raising concerns about the malicious use of fabricated facial content. Existing methods often struggle to generalize to unseen domains due to the diverse nature of facial manipulations. In this paper, we revisit the generation process and identify a universal principle: Deepfake images inherently contain information from both source and target identities, while genuine faces maintain a consistent identity. Building upon this insight, we introduce DiffusionFake, a novel plug-and-play framework that reverses the generative process of face forgeries to enhance the generalization of detection models. DiffusionFake achieves this by injecting the features extracted by the detection model into a frozen pre-trained Stable Diffusion model, compelling it to reconstruct the corresponding target and source images. This guided reconstruction process constrains the detection network to capture the source and target related features to facilitate the reconstruction, thereby learning rich and disentangled representations that are more resilient to unseen forgeries. Extensive experiments demonstrate that DiffusionFake significantly improves cross-domain generalization of various detector architectures without introducing additional parameters during inference. The code are available in https://github.com/skJack/DiffusionFake.git.

Significant progress has been made in the field of Instruction-based Image Editing (IIE). However, evaluating these models poses a significant challenge. A crucial requirement in this field is the establishment of a comprehensive evaluation benchmark for accurately assessing editing results and providing valuable insights for its further development. In response to this need, we propose I2EBench, a comprehensive benchmark designed to automatically evaluate the quality of edited images produced by IIE models from multiple dimensions. I2EBench consists of 2,000+ images for editing, along with 4,000+ corresponding original and diverse instructions. It offers three distinctive characteristics: 1) Comprehensive Evaluation Dimensions: I2EBench comprises 16 evaluation dimensions that cover both high-level and low-level aspects, providing a comprehensive assessment of each IIE model. 2) Human Perception Alignment: To ensure the alignment of our benchmark with human perception, we conducted an extensive user study for each evaluation dimension. 3) Valuable Research Insights: By analyzing the advantages and disadvantages of existing IIE models across the 16 dimensions, we offer valuable research insights to guide future development in the field. We will open-source I2EBench, including all instructions, input images, human annotations, edited images from all evaluated methods, and a simple script for evaluating the results from new IIE models. The code, dataset, and generated images from all IIE models are provided in GitHub: https://github.com/cocoshe/I2EBench.

3D Referring Expression Segmentation (3D-RES) aims to segment 3D objects by correlating referring expressions with point clouds. However, traditional approaches frequently encounter issues like over-segmentation or mis-segmentation, due to insufficient emphasis on spatial information of instances. In this paper, we introduce a Rule-Guided Spatial Awareness Network (RG-SAN) by utilizing solely the spatial information of the target instance for supervision. This approach enables the network to accurately depict the spatial relationships among all entities described in the text, thus enhancing the reasoning capabilities. The RG-SAN consists of the Text-driven Localization Module (TLM) and the Rule-guided Weak Supervision (RWS) strategy. The TLM initially locates all mentioned instances and iteratively refines their positional information. The RWS strategy, acknowledging that only target objects have supervised positional information, employs dependency tree rules to precisely guide the core instance’s positioning. Extensive testing on the ScanRefer benchmark has shown that RG-SAN not only establishes new performance benchmarks, with an mIoU increase of 5.1 points, but also exhibits significant improvements in robustness when processing descriptions with spatial ambiguity. All codes are available at https://github.com/sosppxo/RG-SAN.

In this work, we propose a training-free method to inject visual prompts into Multimodal Large Language Models (MLLMs) through learnable latent variable optimization. We observe that attention, as the core module of MLLMs, connects text prompt tokens and visual tokens, ultimately determining the final results. Our approach involves adjusting visual tokens from the MLP output during inference, controlling the attention response to ensure text prompt tokens attend to visual tokens in referring regions. We optimize a learnable latent variable based on an energy function, enhancing the strength of referring regions in the attention map. This enables detailed region description and reasoning without the need for substantial training costs or model retraining. Our method offers a promising direction for integrating referring abilities into MLLMs, and supports referring with box, mask, scribble and point. The results demonstrate that our method exhibits out-of-domain generalization and interpretability.

One-stage Referring Expression Comprehension (REC) is a task that requires accurate alignment between text descriptions and visual content. In recent years, numerous efforts have been devoted to cross-modal learning for REC, while the influence of other factors in this task still lacks a systematic study. To fill this gap, we conduct an empirical study in this article. Concretely, we ablate 42 candidate designs/settings based on a common REC framework, and these candidates cover the entire process of one-stage REC from network design to model training. Afterwards, we conduct over 100 experimental trials on three REC benchmark datasets. The extensive experimental results reveal the key factors that affect REC performance in addition to multi-modal fusion, e.g., multi-scale features and data augmentation. Based on these findings, we further propose a simple yet strong model called SimREC, which achieves new state-of-the-art performance on these benchmarks. In addition to these progresses, we also find that with much less training overhead and parameters, SimREC can achieve better performance than a set of large-scale pre-trained models, e.g., UNITER and VILLA, portraying the special role of REC in existing V&L research.

Panoptic Narrative Grounding (PNG) is an emerging cross-modal grounding task, which locates the target regions of an image corresponding to the text description. Existing approaches for PNG are mainly based on a two-stage paradigm, which is computationally expensive. In this paper, we propose a one-stage network for real-time PNG, termed End-to-End Panoptic Narrative Grounding network (EPNG), which directly generates masks for referents. Specifically, we propose two innovative designs, i.e., Locality-Perceptive Attention (LPA) and a bidirectional Semantic Alignment Loss (SAL), to properly handle the many-to-many relationship between textual expressions and visual objects. LPA embeds the local spatial priors into attention modeling, i.e., a pixel may belong to multiple masks at different scales, thereby improving segmentation. To help understand the complex semantic relationships, SAL proposes a bidirectional contrastive objective to regularize the semantic consistency inter modalities. Extensive experiments on the PNG benchmark dataset demonstrate the effectiveness and efficiency of our method. Compared to the single-stage baseline, our method achieves a significant improvement of up to 9.4% accuracy. More importantly, our EPNG is 10 times faster than the two-stage model. Meanwhile, the generalization ability of EPNG is also validated by zero-shot experiments on other grounding tasks. The source codes and trained models for all our experiments are publicly available at https://github.com/Mr-Neko/EPNG.git.

Most existing Human-Object Interaction (HOI) Detection methods rely heavily on full annotations with predefined HOI categories, which is limited in diversity and costly to scale further. We aim at advancing zero-shot HOI detection to detect both seen and unseen HOIs simultaneously. The fundamental challenges are to discover potential human-object pairs and identify novel HOI categories. To overcome the above challenges, we propose a novel End-to-end zero-shot HOI Detection (EoID) framework via vision-language knowledge distillation. We first design an Interactive Score module combined with a Two-stage Bipartite Matching algorithm to achieve interaction distinguishment for human-object pairs in an action-agnostic manner. Then we transfer the distribution of action probability from the pretrained vision-language teacher as well as the seen ground truth to the HOI model to attain zero-shot HOI classification. Extensive experiments on HICO-Det dataset demonstrate that our model discovers potential interactive pairs and enables the recognition of unseen HOIs. Finally, our method outperforms the previous SOTA under various zero-shot settings. Moreover, our method is generalizable to large-scale object detection data to further scale up the action sets. The source code is available at: https://github.com/mrwu-mac/EoID.

Building a universal Video-Language model for solving various video understanding tasks (e.g., text-video retrieval, video question answering) is an open challenge to the machine learning field. Towards this goal, most recent works build the model by stacking uni-modal and cross-modal feature encoders and train it with pair-wise contrastive pre-text tasks. Though offering attractive generality, the resulted models have to compromise between efficiency and performance. They mostly adopt different architectures to deal with different downstream tasks. We find this is because the pair-wise training cannot well align and fuse features from different modalities. We then introduce Clover-a Correlated Video-Language pre-training method-towards a universal Video-Language modelfor solving multiple video understanding tasks with neither performance nor efficiency compromise. It improves cross-modal feature alignment and fusion via a novel tri-modal alignment pre-training task. Additionally, we propose to enhance the tri-modal alignment via incorporating learning from semantic masked samples and a new pair-wise ranking loss. Clover establishes new state-of-the-arts on multiple downstream tasks, including three retrieval tasks for both zero-shot and fine-tuning settings, and eight video question answering tasks. Codes and pre-trained models will be released at https://github.com/LeeYN-43/Clover.

Referring Expression Comprehension (REC) is a task of grounding the referent based on an expression, and its development is greatly limited by expensive instance-level annotations. Most existing weakly supervised methods are built based on two-stage detection networks, which are computationally expensive. In this paper, we resort to the efficient one-stage detector and propose a novel weakly supervised model called RefCLIP.Specifically, RefCLIP redefines weakly supervised REC as an anchor-text matching problem, which can avoid the complex post-processing in existing methods. To achieve weakly supervised learning, we introduce anchor-based contrastive loss to optimize Re-fCLlP via numerous anchor-text pairs. Based on RefCLIP, we further propose the first model-agnostic weakly supervised training scheme for existing REC models, where RefCLIP acts as a mature teacher to generate pseudo-labels for teaching common REC models. With our careful designs, this scheme can even help existing REC models achieve better weakly supervised performance than RefCLIP, e.g., TransVg and SimREC. To validate our approaches, we conduct extensive experiments on four REC benchmarks, i.e., RefCOCO, RefCOCO+, RefCOCOg and ReferItGame. Experimental results not only report our significant performance gains over existing weakly supervised models, e.g., +24.87% on RefCOCO, but also show the 5x faster inference speed. Project: https://RefCLIP.github.io.

Referring expression comprehension (REC) often requires a large number of instance-level annotations for fully supervised learning, which are laborious and expensive. In this paper, we present the first attempt of semi-supervised learning for REC and propose a strong baseline method called RefTeacher. Inspired by the recent progress in computer vision, RefTeacher adopts a teacher-student learning paradigm, where the teacher REC network predicts pseudolabels for optimizing the student one. This paradigm allows REC models to exploit massive unlabeled data based on a small fraction of labeled. In particular, we also identify two key challenges in semi-supervised REC, namely, sparse supervision signals and worse pseudo-label noise. To address these issues, we equip RefTeacher with two novel designs called Attention-based Imitation Learning (AIL) and Adaptive Pseudo-label Weighting (APW). AIL can help the student network imitate the recognition behaviors of the teacher, thereby obtaining sufficient supervision signals. APW can help the model adaptively adjust the contributions of pseudo-labels with varying qualities, thus avoiding confirmation bias. To validate RefTeacher, we conduct extensive experiments on three REC benchmark datasets. Experimental results show that RefTeacher obtains obvious gains over the fully supervised methods. More importantly, using only 10% labeled data, our approach allows the model to achieve near 100% fully supervised performance, e.g., only −2.78% on RefCoco. Project: https://refteacher.github.io/.

Text-driven 3D stylization is a complex and crucial task in the fields of computer vision (CV) and computer graphics (CG), aimed at transforming a bare mesh to fit a tar-get text. Prior methods adopt text-independent multilayer perceptrons (MLPs) to predict the attributes of the target mesh with the supervision of CLIP loss. However, such text-independent architecture lacks textual guidance during predicting attributes, thus leading to unsatisfactory stylization and slow convergence. To address these limitations, we present X-Mesh, an innovative text-driven 3D stylization framework that incorporates a novel Text-guided Dynamic Attention Module (TDAM). The TDAM dynamically integrates the guidance of the target text by utilizing text-relevant spatial and channel-wise attentions during vertex feature extraction, resulting in more accurate attribute prediction and faster convergence speed. Furthermore, existing works lack standard benchmarks and automated metrics for evaluation, often relying on subjective and non-reproducible user studies to assess the quality of stylized 3D assets. To overcome this limitation, we introduce a new standard text-mesh benchmark, namely MIT-30, and two automated metrics, which will enable future research to achieve fair and objective comparisons. Our extensive qualitative and quantitative experiments demonstrate that X-Mesh outperforms previous state-of-the-art methods. Our codes and results are available at our project webpage: https://xmu-xiaoma666.github.io/Projects/X-Mesh/

Synthesizing vivid human portraits is a research hot spot in image generation with a wide scope of applications. In addition to fidelity, generation controllability is another key factor that has long plagued its development. To address this issue, existing solutions usually adopt either textual or visual conditions for the target face synthesis, e.g., descriptions or segmentation masks, which still cannot fully control the generation due to the intrinsic shortages of each condition. In this paper, we propose to make use of both types of prior information to facilitate controllable face generation. In particular, we hope to produce coarse-grained information about faces based on the segmentation masks, such as face shapes and poses, and the text description is used to render detailed face attributes, e.g., face color, makeup and gender. More importantly, we hope that the generation can be easily controlled via interactively editing both types of information, making face generation more applicable to real-world applications. To accomplish this target, we propose a novel face generation model termed PixelFace+. In PixelFace+, both the text and mask are encoded as pixel-wise priors, based on which the pixel synthesis process is conducted to produce the expected portraits. Meanwhile, the loss objectives are also carefully designed to make sure that the generated faces are semantically aligned with both text and mask inputs. To validate the proposed PixelFace+, we conducted a comprehensive set of experiments on the widely recognized benchmark called MMCelebA. We not only quantitatively compare PixelFace+ with a bunch of newly proposed Text-to-Face(T2F) generation methods, but also give plenty of qualitative analyses. The experimental results demonstrate that PixelFace+ not only outperforms existing generation methods in both image quality and conditional matching but also shows a much superior controllability of face generation. More importantly, PixelFace+ presents a convenient and interactive way of face generation and manipulation via editing the text and mask inputs. Our SOURCE CODE and DEMO are given in our supplementary materials.

Text-based person retrieval (TPR) is a challenging task that involves retrieving a specific individual based on a textual description. Despite considerable efforts to bridge the gap between vision and language, the significant differences between these modalities continue to pose a challenge. Previous methods have attempted to align text and image samples in a modal-shared space, but they face uncertainties in optimization directions due to the movable features of both modalities and the failure to account for one-to-many relationships of image-text pairs in TPR datasets. To address this issue, we propose an effective bi-directional one-to-many embedding paradigm that offers a clear optimization direction for each sample, thus mitigating the optimization problem. Additionally, this embedding scheme generates multiple features for each sample without introducing trainable parameters, making it easier to align with several positive samples. Based on this paradigm, we propose a novel Bi-directional one-to-many Embedding Alignment (Beat) model to address the TPR task. Our experimental results demonstrate that the proposed Beat model achieves state-of-the-art performance on three popular TPR datasets, including CUHK-PEDES (65.61 R@1), ICFG-PEDES (58.25 R@1), and RSTPReID (48.10 R@1). Furthermore, additional experiments on MS-COCO, CUB, and Flowers datasets further demonstrate the potential of Beat to be applied to other image-text retrieval tasks.

In recent years, 3D representation learning has turned to 2D vision-language pre-trained models to overcome data scarcity challenges. However, existing methods simply transfer 2D alignment strategies, aligning 3D representations with single-view 2D images and coarse-grained parent category text. These approaches introduce information degradation and insufficient synergy issues, leading to performance loss. Information degradation arises from overlooking the fact that a 3D representation should be equivalent to a series of multi-view images and more fine-grained subcategory text. Insufficient synergy neglects the idea that a robust 3D representation should align with the joint vision-language space, rather than independently aligning with each modality. In this paper, we propose a multi-view joint modality modeling approach, termed JM3D, to obtain a unified representation for point cloud, text, and image. Specifically, a novel Structured Multimodal Organizer (SMO) is proposed to address the information degradation issue, which introduces contiguous multi-view images and hierarchical text to enrich the representation of vision and language modalities. A Joint Multi-modal Alignment (JMA) is designed to tackle the insufficient synergy problem, which models the joint modality by incorporating language knowledge into the visual modality. Extensive experiments on ModelNet40 and ScanObjectNN demonstrate the effectiveness of our proposed method, JM3D, which achieves state-of-the-art performance in zero-shot 3D classification. JM3D outperforms ULIP by approximately 4.3% on PointMLP and achieves an improvement of up to 6.5% accuracy on PointNet++ in top-1 accuracy for zero-shot 3D classification on ModelNet40. The source code and trained models for all our experiments are publicly available at https://github.com/Mr-Neko/JM3D.

Despite considerable progress, the advancement of Panoptic Narrative Grounding (PNG) remains hindered by costly annotations. In this paper, we introduce a novel Semi-Supervised Panoptic Narrative Grounding (SS-PNG) learning scheme, capitalizing on a smaller set of labeled image-text pairs and a larger set of unlabeled pairs to achieve competitive performance. Unlike visual segmentation tasks, PNG involves one pixel belonging to multiple open-ended nouns. As a result, existing multi-class based semi-supervised segmentation frameworks cannot be directly applied to this task. To address this challenge, we first develop a novel SS-PNG Network (SS-PNG-NW) tailored to the SS-PNG setting. We thoroughly investigate strategies such as Burn-In and data augmentation to determine the optimal generic configuration for the SS-PNG-NW. Additionally, to tackle the issue of imbalanced pseudo-label quality, we propose a Quality-Based Loss Adjustment (QLA) approach to adjust the semi-supervised objective, resulting in an enhanced SS-PNG-NW+. Employing our proposed QLA, we improve BCE Loss and Dice loss at pixel and mask levels, respectively. We conduct extensive experiments on PNG datasets, with our SS-PNG-NW+ demonstrating promising results comparable to fully-supervised models across all data ratios. Remarkably, our SS-PNG-NW+ outperforms fully-supervised models with only 30% and 50% supervision data, exceeding their performance by 0.8% and 1.1% respectively. This highlights the effectiveness of our proposed SS-PNG-NW+ in overcoming the challenges posed by limited annotations and enhancing the applicability of PNG tasks. The source code is available at https://github.com/nini0919/SSPNG.

Recently, growing interest has been aroused in extending the multimodal capability of large language models (LLMs), e.g., vision-language (VL) learning, which is regarded as the next milestone of artificial general intelligence. However, existing solutions are prohibitively expensive, which not only need to optimize excessive parameters, but also require another large-scale pre-training before VL instruction tuning. In this paper, we propose a novel and affordable solution for the effective VL adaption of LLMs, called Mixture-of-Modality Adaptation (MMA). Instead of using large neural networks to connect the image encoder and LLM, MMA adopts lightweight modules, i.e., adapters, to bridge the gap between LLMs and VL tasks, which also enables the joint optimization of the image and language models. Meanwhile, MMA is also equipped with a routing algorithm to help LLMs achieve an automatic shift between single- and multi-modal instructions without compromising their ability of natural language understanding. To validate MMA, we apply it to a recent LLM called LLaMA and term this formed large vision-language instructed model as LaVIN. To validate MMA and LaVIN, we conduct extensive experiments under two setups, namely multimodal science question answering and multimodal dialogue. The experimental results not only demonstrate the competitive performance and the superior training efficiency of LaVIN than existing multimodal LLMs, but also confirm its great potential as a general-purpose chatbot. More importantly, the actual expenditure of LaVIN is extremely cheap, e.g., only 1.4 training hours with 3.8M trainable parameters, greatly confirming the effectiveness of MMA. Our code is anonymously released at: https://anonymous.4open.science/r/LaVIN--1067.

With ever increasing parameters and computation, vision-language pre-trained (VLP) models exhibit prohibitive expenditure in downstream task adaption. Recent endeavors mainly focus on parameter efficient transfer learning (PETL) for VLP models by only updating a small number of parameters. However, excessive computational overhead still plagues the application of VLPs. In this paper, we aim at parameter and computation efficient transfer learning (PCETL) for VLP models. In particular, PCETL not only needs to limit the number of trainable parameters in VLP models, but also to reduce the computational redundancy during inference, thus enabling a more efficient transfer. To approach this target, we propose a novel dynamic architecture skipping (DAS) approach towards effective PCETL. Instead of directly optimizing the intrinsic architectures of VLP models, DAS first observes the significances of their modules to downstream tasks via a reinforcement learning (RL) based process, and then skips the redundant ones with lightweight networks, i.e. adapters, according to the obtained rewards. In this case, the VLP model can well maintain the scale of trainable parameters while speeding up its inference on downstream tasks. To validate DAS, we apply it to two representative VLP models, namely ViLT and METER, and conduct extensive experiments on a bunch of VL tasks. The experimental results not only show the great advantages of DAS in reducing computational complexity, e.g. -11.97% FLOPs of METER on VQA2.0, but also confirm its competitiveness against existing PETL methods in terms of parameter scale and performance. Our source code is given in our appendix.

Self-attention (SA) based networks have achieved great success in image captioning, constantly dominating the leaderboards of online benchmarks. However, existing SA networks still suffer from distance insensitivity and low-rank bottleneck. In this paper, we aim to optimize SA in terms of two aspects, thereby addressing the above issues. First, we introduce a Distance-sensitive Self-Attention (DSA), which considers the raw geometric distances between query-key pairs in the 2D images during SA modeling. Second, we present a simple yet effective approach, named Multi-branch Self-Attention (MSA) to compensate for the low-rank bottleneck. MSA treats a multi-head self-attention layer as a branch and duplicates it multiple times to increase the expressive power of SA. To validate the effectiveness of the two designs, we apply them to the standard self-attention network, and conduct extensive experiments on the highly competitive MS-COCO dataset. We achieve new state-of-the-art performance on both the local and online test sets,i.e., 135.1% CIDEr on the Karpathy split and 135.4% CIDEr on the official online split.

Attention has become an indispensable component of the models of various multimedia tasks likeImage Captioning(IC) andVisual Question Answering(VQA). However, most existing attention modules are designed for capturing the spatial dependency, and are still insufficient in semantic understanding,e.g., the categories of objects and their attributes, which is also critical for image captioning. To compensate for this defect, we propose a novel attention module termedChannel-wise Attention Block(CAB) to model channel-wise dependency for both visual modality and linguistic modality, thereby improving semantic learning and multi-modal reasoning simultaneously. Specifically, CAB has two novel designs to tackle with the high overhead of channel-wise attention, which are thereduction-reconstruction block structureand thegating-based attention prediction. Based on CAB, we further propose a novelSemantic-enhanced Dual Attention Transformer(termed SDATR), which combines the merits of spatial and channel-wise attentions. To validate SDATR, we conduct extensive experiments on the MS COCO dataset and yield new state-of-the-art performance of 134.5 CIDEr score on COCO Karpathy test split and 136.0 CIDEr score on the official online testing server. To examine the generalization of SDATR, we also apply it to the task of visual question answering, where superior performance gains are also witnessed. The code and models are publicly available athttps://github.com/xmu-xiaoma666/SDATR.

Recent works have validated the benefit of integrating spatial information into deep networks to improve pixel-level prediction tasks such as monocular depth estimation. However, how to efficiently and robustly integrate spatial cues retains as an open problem. In this paper, we introduce the Side Prediction Aggregation (termed SPA) method to enhance the embedding of scene structural information from low-level to high-level layers. To improve the estimation accuracy, the proposed method is further equipped with continuous Spatial Refinement Loss (termed SRL) at multiple resolutions with negligible extra computation. Besides, the proposed sequential network can further perform adversarial learning at multiple resolutions. Such an adversarial refinement strategy greatly improves the accuracy of estimated depth with a little extra computation. Without using any pre-trained models, our network achieves the the-state-of-art accuracy on KITTI, NYUD V2, and Cityscapes datasets, which has achieved real-time depth estimation online.

Referring expression comprehension (REC) is an emerging research topic in computer vision, which refers to the detection of a target region in an image given a test description. Most existing REC methods follow a multistage pipeline, which is computationally expensive and greatly limits the applications of REC. In this article, we propose a one-stage model toward real-time REC, termed real-time global inference network (RealGIN). RealGIN addresses the issues of expression diversity and complexity of REC with two innovative designs: adaptive feature selection (AFS) and Global Attentive ReAsoNing (GARAN). Expression diversity concerns varying expression content, which includes information such as colors, attributes, locations, and fine-grained categories. To address this issue, AFS adaptively fuses features of different semantic levels to tackle the changes in expression content. In contrast, expression complexity concerns the complex relational conditions in expressions that are used to identify the referent. To this end, GARAN uses the textual feature as a pivot to collect expression-aware visual information from all regions and then diffuses this information back to each region, which provides sufficient context for modeling the relational conditions in expressions. On five benchmark datasets, i.e., RefCOCO, RefCOCO+, RefCOCOg, ReferIT, and Flickr30k, the proposed RealGIN outperforms most existing methods and achieves very competitive performances against the most advanced one, i.e., MAttNet. More importantly, under the same hardware, RealGIN can boost the processing speed by 10-20 times over the existing methods.

In this paper, we study teacher-student learning from the perspective of data initialization and propose a novel algorithm called Active Teacher11Source code are available at: https://github.com/HunterJLin/ActiveTeacher for semi-supervised object detection (SSOD). Active Teacher extends the teacher-student framework to an iterative version, where the label set is partially initialized and gradually augmented by evaluating three key factors of unlabeled examples, including difficulty, information and diversity. With this design, Active Teacher can maximize the effect of limited label information while improving the quality of pseudo-labels. To validate our approach, we conduct extensive experiments on the MS-COCO benchmark and compare Active Teacher with a set of recently proposed SSOD methods. The experimental results not only validate the superior performance gain of Active Teacher over the compared methods, but also show that it enables the baseline network, i.e., Faster-RCNN, to achieve 100% supervised performance with much less label expenditure, i.e. 40% labeled examples on MS-COCO. More importantly, we believe that the experimental analyses in this paper can provide useful empirical knowledge for data annotation in practical applications.

Current Image Captioning (IC) methods predict textual words sequentially based on the input visual information from the visual feature extractor and the partially generated sentence information. However, for most cases, the partially generated sentence may dominate the target word prediction due to the insufficiency of visual information, making the generated descriptions irrelevant to the content of the given image. In this paper, we propose a Dual Information Flow Network (DIFNet11Source code is available at: https://github.com/mrwu-mac/DIFNet) to address this issue, which takes segmentation feature as another visual information source to enhance the contribution of visual information for prediction. To maximize the use of two information flows, we also propose an effective feature fusion module termed Iterative Independent Layer Normalization (IILN) which can condense the most relevant inputs while retraining modality-specific information in each flow. Experiments show that our method is able to enhance the dependence of prediction on visual information, making word prediction more focused on the visual content, and thus achieves new state-of-the-art performance on the MSCOCO dataset, e.g., 136.2 CIDEr on COCO Karpathy test split.

Video-text retrieval has been a crucial and fundamental task in multi-modal research. The development of video-text retrieval has been considerably promoted by large-scale multi-modal contrastive pre-training, which primarily focuses on coarse-grained or fine-grained contrast. However, cross-grained contrast, which is the contrast between coarse-grained representations and fine-grained representations, has rarely been explored in prior research. Compared with fine-grained or coarse-grained contrasts, cross-grained contrast calculate the correlation between coarse-grained features and each fine-grained feature, and is able to filter out the unnecessary fine-grained features guided by the coarse-grained feature during similarity calculation, thus improving the accuracy of retrieval. To this end, this paper presents a novel multi-grained contrastive model, namely X-CLIP, for video-text retrieval. However, another challenge lies in the similarity aggregation problem, which aims to aggregate fine-grained and cross-grained similarity matrices to instance-level similarity. To address this challenge, we propose the Attention Over Similarity Matrix (AOSM) module to make the model focus on the contrast between essential frames and words, thus lowering the impact of unnecessary frames and words on retrieval results. With multi-grained contrast and the proposed AOSM module, X-CLIP achieves outstanding performance on five widely-used video-text retrieval datasets, including MSR-VTT (49.3 [email protected]), MSVD (50.4 [email protected]), LSMDC (26.1 [email protected]), DiDeMo (47.8 [email protected]) and ActivityNet (46.2 [email protected]).

Text-to-face (T2F) generation is an emerging research hot spot in multimedia, which aims to synthesize vivid portraits based on the given descriptions. Its main challenge lies in the accurate alignments from texts to image pixels, which pose a high demand in generation fidelity. We define T2F as a pixel synthesis problem conditioned on the texts and propose a novel dynamic pixel synthesis network, PixelFace, for end-to-end T2F generation in this paper. To fully exploit the prior knowledge for T2F synthesis, we propose a novel dynamic parameter generation module, which transforms text features into dynamic knowledge embeddings for end-to-end pixel regression. These knowledge embeddings are example-dependent and spatially related to image pixels, based on which PixelFace can exploit the text priors for high-quality text-guided face generation. To validate the proposed PixelFace, we conduct extensive experiments on the MMCelebA, and compare PixelFace with a set of state-of-the-art methods in T2F and T2I generations, e.g., StyleCLIP and TediGAN. The experimental results not only show the greater performance of PixelFace than the compared methods but also validates its merits over existing T2F methods in both text-image matching and inference speed. Codes will be released at: \textcolormagenta \urlhttps://github.com/pengjunn/PixelFace .

Text-to-face (T2F) generation is an emerging research hot spot in multimedia, and its main challenge lies in the high fidelity requirement of generated portraits. Many existing works resort to exploring the latent space in a pre-trained generator, e.g., StyleGAN, which has obvious shortcomings in efficiency and generalization ability. In this paper, we propose a generative network for open-ended text-to-face generation, which is termed OpenFaceGAN. Differing from existing StyleGAN-based methods, OpenFaceGAN constructs an effective multi-modal latent space that directly converts the natural language description into a face. This mapping paradigm can fit the real data distribution well and make the model capable of open-ended and even zero-shot T2F generation. Our method improves the inference speed by an order of magnitude, e.g., 294 times than TediGAN. Based on OpenFaceGAN, we further explore text-guided face manipulation (editing). In particular, we propose a parameterized module, OpenEditor, to automatically disentangle the target latent code and update the original style information. OpenEditor also makes OpenFaceGAN directly applicable for most manipulation instructions without example-dependent searches or optimizations, greatly improving the efficiency of face manipulation. We conduct extensive experiments on two benchmark datasets namely Multi-Modal CelebA-HQ and Face2Text-v1.0. The experimental results not only show the superior performance of OpenFaceGAN to the existing T2F methods in both image quality and image-text matching degree but also greatly confirm its outstanding ability in the zero-shot generation. Codes will be released at: \textcolormagenta \urlhttps://github.com/pengjunn/OpenFace

Deep neural networks often suffer from poor generalization caused by complex and non-convex loss landscapes. One of the popular solutions is Sharpness-Aware Minimization (SAM), which smooths the loss landscape via minimizing the maximized change of training loss when adding a perturbation to the weight. However, we find the indiscriminate perturbation of SAM on all parameters is suboptimal, which also results in excessive computation,~\emph{i.e.}, double the overhead of common optimizers like Stochastic Gradient Descent~(SGD). In this paper, we propose an efficient and effective training scheme coined as Sparse SAM (SSAM), which achieves sparse perturbation by a binary mask. To obtain the sparse mask, we provide two solutions which are based on Fisher information and dynamic sparse training, respectively. In addition, we theoretically prove that SSAM can converge at the same rate as SAM,~\emph{i.e.}, $O(\log T/\sqrt{T})$. Sparse SAM not only has the potential for training acceleration but also smooths the loss landscape effectively. Extensive experimental results on CIFAR10, CIFAR100, and ImageNet-1K confirm the superior efficiency of our method to SAM, and the performance is preserved or even better with a perturbation of merely 50\% sparsity. Code is available at \url{https://github.com/Mi-Peng/Sparse-Sharpness-Aware-Minimization}.

Online image hashing has received increasing research attention recently, which processes large-scale data in a streaming fashion to update the hash functions on-the-fly. To this end, most existing works exploit this problem under a supervised setting, i.e., using class labels to boost the hashing performance, which suffers from the defects in both adaptivity and efficiency: First, large amounts of training batches are required to learn up-to-date hash functions, which leads to poor online adaptivity. Second, the training is time-consuming, which contradicts with the core need of online learning. In this paper, a novel supervised online hashing scheme, termed Fast Class-wise Updating for Online Hashing (FCOH), is proposed to address the above two challenges by introducing a novel and efficient inner product operation. To achieve fast online adaptivity, a class-wise updating method is developed to decompose the binary code learning and alternatively renew the hash functions in a class-wise fashion, which well addresses the burden on large amounts of training batches. Quantitatively, such a decomposition further leads to at least 75 percent storage saving. To further achieve online efficiency, we propose a semi-relaxation optimization, which accelerates the online training by treating different binary constraints independently. Without additional constraints and variables, the time complexity is significantly reduced. Such a scheme is also quantitatively shown to well preserve past information during updating hashing functions. We have quantitatively demonstrated that the collective effort of class-wise updating and semi-relaxation optimization provides a superior performance comparing to various state-of-the-art methods, which is verified through extensive experiments on three widely-used datasets.

Visual Question Answering (VQA) has attracted extensive research focus recently. Along with the ever-increasing data scale and model complexity, the enormous training cost has become an emerging challenge for VQA. In this article, we show such a massive training cost is indeed plague. In contrast, a fine-grained design of the learning paradigm can be extremely beneficial in terms of both training efficiency and model accuracy. In particular, we argue that there exist two essential and unexplored issues in the existing VQA training paradigm that randomly samples data in each epoch, namely, the "difficulty diversity" and the "label redundancy". Concretely, "difficulty diversity" refers to the varying difficulty levels of different question types, while "label redundancy" refers to the redundant and noisy labels contained in individual question type. To tackle these two issues, in this article we propose a fine-grained VQA learning paradigm with an actor-critic based learning agent, termed FG-A1C. Instead of using all training data from scratch, FG-A1C includes a learning agent that adaptively and intelligently schedules the most difficult question types in each training epoch. Subsequently, two curriculum learning based schemes are further designed to identify the most useful data to be learned within each inidividual question type. We conduct extensive experiments on the VQA2.0 and VQA-CP v2 datasets, which demonstrate the significant benefits of our approach. For instance, on VQA-CP v2, with less than 75 percent of the training data, our learning paradigms can help the model achieves better performance than using the whole dataset. Meanwhile, we also shows the effectivenesss of our method in guiding data labeling. Finally, the proposed paradigm can be seamlessly integrated with any cutting-edge VQA models, without modifying their structures.

Despite considerable progress, image captioning still suffers from the huge difference in quality between easy and hard examples, which is left unexploited in existing methods. To address this issue, we explore the hard example mining in image captioning, and propose a simple yet effective mechanism to instruct the model to pay more attention to hard examples, thereby improving the performance in both general and complex scenarios. We first propose a novel learning strategy, termed Metric-oriented Focal Mechanism (MFM), for hard example mining in image captioning. Differing from the existing strategies for classification tasks, MFM can adopt the generative metrics of image captioning to measure the difficulties of examples, and then up-weight the rewards of hard examples during training. To make MFM applicable to different datasets without tedious parameter tuning, we further introduce an adaptive reward metric called Effective CIDEr (ECIDEr), which considers the data distribution of easy and hard examples during reward estimation. Extensive experiments are conducted on the MS COCO benchmark, and the results show that while maintaining the performance on simple examples, MFM can significantly improve the quality of captions for hard examples. The ECIDEr-based MFM is equipped on the current SOTA method, e.g., DLCT (Luo et al., 2021), which outperforms all existing methods and achieves new state-of-the-art performance on both the off-line and on- line testing, i.e., 134.3 CIDEr for the off-line testing and 136.1 for the on- line testing of MSCOCO. To validate the generalization ability of ECIDEr-based MFM, we also apply it to another dataset, namely Flickr30k, and superior performance gains can also be obtained.

Despite the exciting performance, Transformer is criticized for its excessive parameters and computation cost. However, compressing Transformer remains as an open problem due to its internal complexity of the layer designs, i.e., Multi-Head Attention (MHA) and Feed-Forward Network (FFN). To address this issue, we introduce Group-wise Transformation towards a universal yet lightweight Transformer for vision-and-language tasks, termed as LW-Transformer. LW-Transformer applies Group-wise Transformation to reduce both the parameters and computations of Transformer, while also preserving its two main properties, i.e., the efficient attention modeling on diverse subspaces of MHA, and the expanding-scaling feature transformation of FFN. We apply LW-Transformer to a set of Transformer-based networks, and quantitatively measure them on three vision-and-language tasks and six benchmark datasets. Experimental results show that while saving a large number of parameters and computations, LW-Transformer achieves very competitive performance against the original Transformer networks for vision-and-language tasks. To examine the generalization ability, we apply LW-Transformer to the task of image classification, and build its network based on a recently proposed image Transformer called Swin-Transformer, where the effectiveness can be also confirmed.

Text-to-Image (T2I) synthesis is a challenging task that aims to convert natural language descriptions to real images. It remains an open problem mainly due to the diversity of text descriptions, which poses a huge obstacle in generating vivid and relevant images. Moreover, the existing evaluation metrics in T2I synthesis are mainly used to evaluate the visual quality of the generated images, while the semantic consistency between the two modalities is often ignored. To address these issues, we present a novelKnowledge-Driven Generative Adversarial Network, termed KD-GAN, and a new evaluation system, namedPseudo Turing Test(PTT for short). Concretely, KD-GAN takes a further step in imitating the behavior of human painting,i.e., drawing an image according to reference knowledge. The introduction of reference knowledge in KD-GAN not only improves the quality of the generated images but also enhances the semantic consistency between them and the input texts. In addition, KD-GAN can also greatly avoid some flaws against common sense during image generation,e.g., skiing in the blue sky. The proposed PTT is an important supplement to the existing evaluation system of T2I synthesis. It includes a set of pseudo-experts of different multimedia tasks to evaluate the semantic consistency between the given texts and the generated images. To validate the proposed KD-GAN, we conducted extensive experiments on two benchmark datasets,i.e., Caltech-UCSD Birds (CUB), and MS-COCO (COCO). The experimental results demonstrate that KD-GAN outperforms state-of-the-art methods on IS, FID, and the proposed PTT metrics.11The codes of KD-GAN are at [Online]. Available:https://github.com/pengjunn/KD-GANand the codes and models of PTT are at [Online]. Available:https://github.com/pengjunn/PTT.

Transformer-based architectures have shown great success in image captioning, where object regions are encoded and then attended into the vectorial representations to guide the caption decoding. However, such vectorial representations only contain region-level information without considering the global information reflecting the entire image, which fails to expand the capability of complex multi-modal reasoning in image captioning. In this paper, we introduce a Global Enhanced Transformer (termed GET) to enable the extraction of a more comprehensive global representation, and then adaptively guide the decoder to generate high-quality captions. In GET, a Global Enhanced Encoder is designed for the embedding of the global feature, and a Global Adaptive Decoder are designed for the guidance of the caption generation. The former models intra- and inter-layer global representation by taking advantage of the proposed Global Enhanced Attention and a layer-wise fusion module. The latter contains a Global Adaptive Controller that can adaptively fuse the global information into the decoder to guide the caption generation. Extensive experiments on MS COCO dataset demonstrate the superiority of our GET over many state-of-the-arts.

Descriptive region features extracted by object detection networks have played an important role in the recent advancements of image captioning. However, they are still criticized for the lack of contextual information and fine-grained details, which in contrast are the merits of traditional grid features. In this paper, we introduce a novel Dual-Level Collaborative Transformer (DLCT) network to realize the complementary advantages of the two features. Concretely, in DLCT, these two features are first processed by a novel Dual-way Self Attenion (DWSA) to mine their intrinsic properties, where a Comprehensive Relation Attention component is also introduced to embed the geometric information. In addition, we propose a Locality-Constrained Cross Attention module to address the semantic noises caused by the direct fusion of these two features, where a geometric alignment graph is constructed to accurately align and reinforce region and grid features. To validate our model, we conduct extensive experiments on the highly competitive MS-COCO dataset, and achieve new state-of-the-art performance on both local and online test sets, i.e., 133.8% CIDEr on Karpathy split and 135.4% CIDEr on the official split.

Recent progress on visual question answering has explored the merits of grid features for vision language tasks. Meanwhile, transformer-based models have shown remarkable performance in various sequence prediction problems. However, the spatial information loss of grid features caused by flattening operation, as well as the defect of the transformer model in distinguishing visual words and non visual words, are still left unexplored. In this paper, we first propose Grid-Augmented (GA) module, in which relative geometry features between grids are incorporated to enhance visual representations. Then, we build a BERT-based language model to extract language context and propose Adaptive-Attention (AA) module on top of a transformer decoder to adaptively measure the contribution of visual and language cues before making decisions for word prediction. To prove the generality of our proposals, we apply the two modules to the vanilla transformer model to build our Relationship-Sensitive Transformer (RSTNet) for image captioning task. The proposed model is tested on the MSCOCO benchmark, where it achieves new state-of-art results on both the Karpathy test split and the online test server. Source code is available at GitHub1.

Due to the superior ability of global dependency modeling, Transformer and its variants have become the primary choice of many vision-and-language tasks. However, in tasks like Visual Question Answering (VQA) and Referring Expression Comprehension (REC), the multimodal prediction often requires visual information from macro- to micro-views. Therefore, how to dynamically schedule the global and local dependency modeling in Transformer has become an emerging issue. In this paper, we propose an example-dependent routing scheme called TRAnsformer Routing (TRAR) to address this issue1. Specifically, in TRAR, each visual Transformer layer is equipped with a routing module with different attention spans. The model can dynamically select the corresponding attentions based on the output of the previous inference step, so as to formulate the optimal routing path for each example. Notably, with careful designs, TRAR can reduce the additional computation and memory overhead to almost negligible. To validate TRAR, we conduct extensive experiments on five benchmark datasets of VQA and REC, and achieve superior performance gains than the standard Transformers and a bunch of state-of-the-art methods.

Automated machine learning (AutoML) has achieved remarkable progress on various tasks, which is attributed to its minimal involvement of manual feature and model designs. However, most of existing AutoML pipelines only touch parts of the full machine learning pipeline, e.g., neural architecture search or optimizer selection. This leaves potentially important components such as data cleaning and model ensemble out of the optimization, and still results in considerable human involvement and suboptimal performance. The main challenges lie in the huge search space assembling all possibilities over all components, as well as the generalization ability over different tasks like image, text, and tabular etc. In this paper, we present a first-of-its-kind fully AutoML pipeline, to comprehensively automate data preprocessing, feature engineering, model generation/selection/training and ensemble for an arbitrary dataset and evaluation metric. Our innovation lies in the comprehensive scope of a learning pipeline, with a novel "life-long" knowledge anchor design to fundamentally accelerate the search over the full search space. Such knowledge anchors record detailed information of pipelines and integrates them with an evolutionary algorithm for joint optimization across components. Experiments demonstrate that the result pipeline achieves state-of-the-art performance on multiple datasets and modalities. Specifically, the proposed framework was extensively evaluated in the NeurIPS 2019 AutoDL challenge, and won the only champion with a significant gap against other approaches, on all the image, video, speech, text and tabular tracks.

In this paper, we propose a novel deep framework for part-level semantic parsing of freehand sketches, which makes three main contributions that are experimentally shown to have substantial practical merit. First, we propose a homogeneous transformation method to address the problem of domain adaptation. For the task of sketch parsing, there is no available data of labeled freehand sketches that can be directly used for model training. An alternative solution is to learn from datasets of real image parsing, while the domain adaptation is an inevitable problem. Unlike existing methods that utilize the edge maps of real images to approximate freehand sketches, the proposed homogeneous transformation method transforms the data from domains of real images and freehand sketches into a homogeneous space to minimize the semantic gap. Second, we design a soft-weighted loss function as guidance for the training process, which gives attention to both the ambiguous label boundary and class imbalance. Third, we present a staged learning strategy to improve the parsing performance of the trained model, which takes advantage of the shared information and specific characteristic from different sketch categories. Extensive experimental results demonstrate the effectiveness of the above three methods. Specifically, to evaluate the generalization ability of our homogeneous transformation method, additional experiments for the task of sketch-based image retrieval are conducted on the QMUL FG-SBIR dataset. Finally, by integrating the proposed three methods into a unified framework of deep semantic sketch parsing (DeepSSP), we achieve the state-of-the-art on the public SketchParse dataset.

Deep hashing methods have been proved to be effective and efficient for large-scale Web media search. The success of these data-driven methods largely depends on collecting sufficient labeled data, which is usually a crucial limitation in practical cases. The current solutions to this issue utilize Generative Adversarial Network (GAN) to augment data in semi-supervised learning. However, existing GAN-based methods treat image generations and hashing learning as two isolated processes, leading to generation ineffectiveness. Besides, most works fail to exploit the semantic information in unlabeled data. In this paper, we propose a novel Semi-supervised Self-pace Adversarial Hashing method, named SSAH to solve the above problems in a unified framework. The SSAH method consists of an adversarial network (A-Net) and a hashing network (H-Net). To improve the quality of generative images, first, the A-Net learns hard samples with multi-scale occlusions and multi-angle rotated deformations which compete against the learning of accurate hashing codes. Second, we design a novel self-paced hard generation policy to gradually increase the hashing difficulty of generated samples. To make use of the semantic information in unlabeled ones, we propose a semi-supervised consistent loss. The experimental results show that our method can significantly improve state-of-the-art models on both the widely-used hashing datasets and fine-grained datasets.

Referring expression comprehension (REC) and segmentation (RES) are two highly-related tasks, which both aim at identifying the referent according to a natural language expression. In this paper, we propose a novel Multi-task Collaborative Network (MCN) to achieve a joint learning of REC and RES for the first time. In MCN, RES can help REC to achieve better language-vision alignment, while REC can help RES to better locate the referent. In addition, we address a key challenge in this multi-task setup, i.e., the prediction conflict, with two innovative designs namely, Consistency Energy Maximization (CEM) and Adaptive Soft Non-Located Suppression (ASNLS). Specifically, CEM enables REC and RES to focus on similar visual regions by maximizing the consistency energy between two tasks. ASNLS supresses the response of unrelated regions in RES based on the prediction of REC. To validate our model, we conduct extensive experiments on three benchmark datasets of REC and RES, i.e., RefCOCO, RefCOCO+ and RefCOCOg. The experimental results report the significant performance gains of MCN over all existing methods, i.e., up to +7.13% for REC and +11.50% for RES over SOTA, which well confirm the validity of our model for joint REC and RES learning.

In this paper, we investigate the fragility of deep image captioning models against adversarial attacks. Different from existing works that generate common words and concepts, we focus on the adversarial attacks towards controllable image captioning, i.e., removing target words from captions by imposing adversarial noises to images while maintaining the captioning accuracy for the remaining visual content. We name this new task as Masked Image Captioning (MIC), which is expected to be training and labeling free for end-to-end captioning models. Meanwhile, we propose a novel adversarial learning approach for this new task, termed Show, Mask, and Tell (SMT), which crafts adversarial examples to mask the target concepts via minimizing an objective loss while training the noise generator. Concretely, three novel designs are introduced in this loss, i.e., word removal regularization, captioning accuracy regularization, and noise filtering regularization. For quantitative validation, we propose a benchmark dataset for MIC based on the MS COCO dataset, together with a new evaluation metric called Attack Quality. Experimental results show that the proposed approach achieves successful attacks by removing 93.8% and 91.9% target words while maintaining 97.3% and 97.4% accuracies on two cutting-edge captioning models, respectively.

Referring expression segmentation (RES) aims to segment the target instance in a given image according to a natural language expression. Its main challenge lies in how to quickly and accurately align the text expression to the referred visual instances. In this paper, we focus on addressing this issue by proposing a Cascade Grouped Attention Network (CGAN) with two innovative designs: Cascade Grouped Attention (CGA) and Instance-level Attention (ILA) loss. Specifically, CGA is used to perform step-wise reasoning over the entire image to perceive the differences between instances accurately yet efficiently, so as to identify the referent. ILA loss is further embedded into each step of CGA to directly supervise the attention modeling, which improves the alignments between the text expression and the visual instances. Through these two novel designs, CGAN can achieve the high efficiency of one-stage RES while possessing a strong reasoning ability comparable to the two-stage methods. To validate our model, we conduct extensive experiments on three RES benchmark datasets and achieve significant performance gains over existing one-stage and multi-stage models

Modeling human visual attention mechanism is a fundamental problem for the understanding of human vision, which has also been demonstrated as an important module for various multimedia applications such as image captioning and visual question answering. In this paper, we propose a new probabilistic framework for attention, and introduce the concept ofmode to model the flexibility and adaptability of attention modulation in complex environments. We characterize the correlations between the visual input, the activated mode, the saliency and the spatial allocation of attention via a graphical model representation, based on which we explore the lingual guidance from captioning data for the implementation of a mode-sensitive attention (MSA) model. The proposed framework explicitly justifies the usage of center bias for fixation prediction and can convert an arbitrary learning-based backbone attention model to a more robust multi-mode version. Experimental results on the York120, MIT1003 and PASCAL datasets demonstrate the effectiveness of the proposed method.

In this paper, we propose a cross-modal network architecture search (NAS) algorithm for VQA, termed as k-Armed Bandit based NAS (KAB-NAS). KAB-NAS regards the design of each layer as a k-armed bandit problem and updates the preference of each candidate via numerous samplings in a single-shot search framework. To establish an effective search space, we further propose a new architecture termed Automatic Graph Attention Network (AGAN), and extend the popular self-attention layer with three graph structures, denoted as dense-graph, co-graph and separate-graph.These graph layers are used to form the direction of information propagation in the graph network, and their optimal combinations are searched by KAB-NAS. To evaluate KAB-NAS and AGAN, we conduct extensive experiments on two VQA benchmark datasets, i.e., VQA2.0 and GQA, and also test AGAN with the popular BERT-style pre-training. The experimental results show that with the help of KAB-NAS, AGAN can achieve the state-of-the-art performance on both benchmark datasets with much fewer parameters and computations.

In this paper, we address the problem of monocular depth estimation when only a limited number of training image-depth pairs are available. To achieve a high regression accuracy, the state-of-the-art estimation methods rely on CNNs trained with a large number of image-depth pairs, which are prohibitively costly or even infeasible to acquire. Aiming to break the curse of such expensive data collections, we propose a semi-supervised adversarial learning framework that only utilizes a small number of image-depth pairs in conjunction with a large number of easily-available monocular images to achieve high performance. In particular, we use one generator to regress the depth and two discriminators to evaluate the predicted depth, i.e., one inspects the image-depth pair while the other inspects the depth channel alone. These two discriminators provide their feedbacks to the generator as the loss to generate more realistic and accurate depth predictions. Experiments show that the proposed approach can (1) improve most state-of-the-art models on the NYUD v2 dataset by effectively leveraging additional unlabeled data sources; (2) reach state-of-the-art accuracy when the training set is small, e.g., on the Make3D dataset; (3) adapt well to an unseen new dataset (Make3D in our case) after training on an annotated dataset (KITTI in our case).

In recent years, hashing methods have been proved to be effective and efficient for large-scale Web media search. However, the existing general hashing methods have limited discriminative power for describing fine-grained objects that share similar overall appearance but have a subtle difference. To solve this problem, we for the first time introduce the attention mechanism to the learning of fine-grained hashing codes. Specifically, we propose a novel deep hashing model, named deep saliency hashing (DSaH), which automatically mines salient regions and learns semantic-preserving hashing codes simultaneously. DSaH is a two-step end-to-end model consisting of an attention network and a hashing network. Our loss function contains three basic components, including the semantic loss, the saliency loss, and the quantization loss. As the core of DSaH, the saliency loss guides the attention network to mine discriminative regions from pairs of images.We conduct extensive experiments on both fine-grained and general retrieval datasets for performance evaluation. Experimental results on fine-grained datasets, including Oxford Flowers, Stanford Dogs, and CUB Birds demonstrate that our DSaH performs the best for the fine-grained retrieval task and beats the strongest competitor (DTQ) by approximately 10% on both Stanford Dogs and CUB Birds. DSaH is also comparable to several state-of-the-art hashing methods on CIFAR-10 and NUS-WIDE.

Online image hashing aims to update hash functions on-the-fly along with newly arriving data streams, which has found broad applications in computer vision and beyond. To this end, most existing methods update hash functions simply using discrete labels or pairwise similarity to explore intra-class relationships, which, however, often deteriorates search performance when facing a domain gap or semantic shift. One reason is that they ignore the particular semantic relationships among different classes, which should be taken into account in updating hash functions. Besides, the common characteristics between the label vectors (can be regarded as a sort of binary codes) and to-be-learned binary hash codes have left unexploited. In this paper, we present a novel online hashing method, termed Similarity Preserving Linkage Hashing (SPLH), which not only utilizes pairwise similarity to learn the intra-class relationships, but also fully exploits a latent linkage space to capture the inter-class relationships and the common characteristics between label vectors and to-be-learned hash codes. Specifically, SPLH first maps the independent discrete label vectors and binary hash codes into a linkage space, through which the relative semantic distance between data points can be assessed precisely. As a result, the pairwise similarities within the newly arriving data stream are exploited to learn the latent semantic space to benefit binary code learning. To learn the model parameters effectively, we further propose an alternating optimization algorithm. Extensive experiments conducted on three widely-used datasets demonstrate the superior performance of SPLH over several state-of-the-art online hashing methods.

When facing large-scale image datasets, online hashing serves as a promising solution for online retrieval and prediction tasks. It encodes the online streaming data into compact binary codes, and simultaneously updates the hash functions to renew codes of the existing dataset. To this end, the existing methods update hash functions solely based on the new data batch, without investigating the correlation between such new data and the existing dataset. In addition, existing works update the hash functions using a relaxation process in its corresponding approximated continuous space. And it remains as an open problem to directly apply discrete optimizations in online hashing. In this paper, we propose a novel supervised online hashing method, termed Balanced Similarity for Online Discrete Hashing (BSODH), to solve the above problems in a unified framework. BSODH employs a well-designed hashing algorithm to preserve the similarity between the streaming data and the existing dataset via an asymmetric graph regularization. We further identify the “data-imbalance” problem brought by the constructed asymmetric graph, which restricts the application of discrete optimization in our problem. Therefore, a novel balanced similarity is further proposed, which uses two equilibrium factors to balance the similar and dissimilar weights and eventually enables the usage of discrete optimizations. Extensive experiments conducted on three widely-used benchmarks demonstrate the advantages of the proposed method over the stateof-the-art methods.

Recent advances on fine-grained image retrieval prefer learning convolutional neural network (CNN) with specific fullyconnect layer designed loss function for discriminative feature representation. Essentially, such loss should establish a robust metric to efficiently distinguish high-dimensional features within and outside fine-grained categories. To this end, the existing loss functions are defected in two aspects: (a) The feature relationship is encoded inside the training batch. Such a local scope leads to low accuracy. (b) The error is established by the mean square, which needs pairwise distance computation in training set and results in low efficiency. In this paper, we propose a novel metric learning scheme, termed Normalize-Scale Layer and Decorrelated Global Centralized Ranking Loss, which achieves extremely efficient and discriminative learning, i.e., 5× speedup over triplet loss and 12% recall boost on CARS196. Our method originates from the classic softmax loss, which has a global structure but does not directly optimize the distance metric as well as the inter/intra class distance. We tackle this issue through a hypersphere layer and a global centralized ranking loss with a pairwise decorrelated learning. In particular, we first propose a Normalize-Scale Layer to eliminate the gap between metric distance (for measuring distance in retrieval) and dot product (for dimension reduction in classification). Second, the relationship between features is encoded under a global centralized ranking loss, which targets at optimizing metric distance globally and accelerating learning procedure. Finally, the centers are further decorrelated by Gram-Schmidt process, leading to extreme efficiency (with 20 epochs in training procedure) and discriminability in feature learning. We have conducted quantitative evaluations on two fine-grained retrieval benchmark. The superior performance demonstrates the merits of the proposed approach over the state-of-the-arts.

In this paper, we uncover the issue of knowledge inertia in visual question answering (VQA), which commonly exists in most VQA models and forces the models to mainly rely on the question content to “guess” answer, without regard to the visual information. Such an issue not only impairs the performance of VQA models, but also greatly reduces the credibility of the answer prediction. To this end, simply highlighting the visual features in the model is undoable, since the prediction is built upon the joint modeling of two modalities and largely influenced by the data distribution. In this paper, we propose a Pairwise Inconformity Learning (PIL) to tackle the issue of knowledge inertia. In particular, PIL takes full advantage of the similar image pairs with diverse answers to an identical question provided in VQA2.0 dataset. It builds a multi-modal embedding space to project pos./neg. feature pairs, upon which word vectors of answers are modeled as anchors. By doing so, PIL strengthens the importance of visual features in prediction with a novel dynamic-margin based triplet loss that efficiently increases the semantic discrepancies between pos./neg. image pairs. To verify the proposed PIL, we plug it on a baseline VQA model as well as a set of recent VQA models, and conduct extensive experiments on two benchmark datasets, i.e., VQA1.0 and VQA2.0. Experimental results show that PIL can boost the accuracy of the existing VQA models (1.56%-2.93% gain) with a negligible increase in parameters (0.85%-5.4% parameters). Qualitative results also reveal the elimination of knowledge inertia in the existing VQA models after implementing our PIL.

In visual question answering (VQA), recent advances have well advocated the use of attention mechanism to precisely link the question to the potential answer areas. As the difficulty of the question increases, more VQA models adopt multiple attention layers to capture the deeper visual-linguistic correlation. But a negative consequence is the explosion of parameters, which makes the model vulnerable to over-fitting, especially when limited training examples are given. In this paper, we propose an extremely compact alternative to this static multi-layer architecture towards accurate yet efficient attention modeling, termed as Dynamic Capsule Attention (CapsAtt). Inspired by the recent work of Capsule Network, CapsAtt treats visual features as capsules and obtains the attention output via dynamic routing, which updates the attention weights by calculating coupling coefficients between the underlying and output capsules. Meanwhile, CapsAtt also discards redundant projection matrices to make the model much more compact. We quantify CapsAtt on three benchmark VQA datasets, i.e., COCO-QA, VQA1.0 and VQA2.0. Compared to the traditional multi-layer attention model, CapsAtt achieves significant improvements of up to 4.1%, 5.2% and 2.2% on three datasets, respectively. Moreover, with much fewer parameters, our approach also yields competitive results compared to the latest VQA models. To further verify the generalization ability of CapsAtt, we also deploy it on another challenging multi-modal task of image captioning, where state-of-the-art performance is achieved with a simple network structure.

The cross-modality topic detection in social media retains as an open problem mainly due to the difficulty of dealing with modality independence and modality missing. In this paper, we present a novel Deep Topical Correlation Analysis (DTCA) approach, which achieves robust and accurate topic detecting for micro-blogs and handles aforementioned challenges simultaneously. In particular, bidirectional recurrent neural networks and convolutional neural networks are used to learn deep textual and visual features, respectively. Then a Canonical Correlation Analysis based fusion scheme is proposed, which has two innovations to deal with both two problems mentioned above. We further release a large-scale cross-modal twitter dataset for topic detection. Extensive and quantitative evaluations are conducted with comparisons to several state-of-the-arts and alternative approaches on this dataset. Significant performance gains are reported to demonstrate the merits of proposed approach.

Recovering the 3D representation of an object from single-view or multi-view RGB images by deep neural networks has attracted increasing attention in the past few years. Several mainstream works (e.g., 3D-R2N2) use recurrent neural networks (RNNs) to fuse multiple feature maps extracted from input images sequentially. However, when given the same set of input images with different orders, RNN-based approaches are unable to produce consistent reconstruction results. Moreover, due to long-term memory loss, RNNs cannot fully exploit input images to refine reconstruction results. To solve these problems, we propose a novel framework for single-view and multi-view 3D reconstruction, named Pix2Vox. By using a well-designed encoder-decoder, it generates a coarse 3D volume from each input image. Then, a context-aware fusion module is introduced to adaptively select high-quality reconstructions for each part (e.g., table legs) from different coarse 3D volumes to obtain a fused 3D volume. Finally, a refiner further refines the fused 3D volume to generate the final output. Experimental results on the ShapeNet and Pix3D benchmarks indicate that the proposed Pix2Vox outperforms state-of-the-arts by a large margin. Furthermore, the proposed method is 24 times faster than 3D-R2N2 in terms of backward inference time. The experiments on ShapeNet unseen 3D categories have shown the superior generalization abilities of our method.

Block-based hybrid coding is widely used in video compression. As the bit rate decreases, the quantization becomes lossy resulting in unacceptable blocking artifacts. Although advanced video encoder with loop filter has achieved promising results, the problem still remains unsolved. Most of the previous solutions regard a video sequence as a group of independent frames, without considering their temporal relationships. To address the above issue, this paper designs a temporal boosting residual network aiming to integrate the temporal information and the structural information. And the information is represented as the residual values of previous frames which are extracted by the deep residual network. The entire framework is a cascading architecture that implies a coarse-to-fine processing. The proposed solution does not modify any module of the codec, it only takes the lossy frames as input, and outputs the enhanced frames, which is a typical end-to-end mapping. Experimental results validate that our framework has achieved 0.6-1.0 dB improvement on average based on HEVC software and outperformed the state-of-the-art methods in both objective and perceptual quality.

Deep convolutional neural networks (DCNN) with manifold embedding have achieved considerable attention in computer vision. However, prior arts are usually based on the neighborhood-based graph modeling only the pairwise relationship between two samples, which fail to fully capture intra-class variations and thus suffer from severe performance loss for noisy data. While such intra-class variations can be well captured via sophisticated hypergraph structure, we are motivated and lead a hypergraph induced Convolutional Manifold Network (H-CMN) to significantly improve the representation capacity of DCNN for the complex data. Specifically, two innovative designs are provides: 1) our manifold preserving method is implemented based on a mini-batch, which can be efficiently plugged into the existing DCNN training pipelines and be scalable for large datasets; 2) a robust hypergraph is built for each mini-batch, which not only offers a strong robustness against typical noise, but also captures the variances from multiple features. Extensive experiments on the image classification task on large benchmarking datasets demonstrate that our model achieves much better performance than the state-of-the-art

Face anti-spoofing detection is critical to guarantee the security of biometric face recognition systems. Despite extensive advances in facial anti-spoofing based on single-model image, little work has been devoted to multi-modal anti-spoofing, which is however widely encountered in real-world scenarios. Following the recent progress, this paper mainly focuses on multi-modal face anti-spoofing and aims to solve the following two challenges: (1) how to effectively fuse multi-modal information; and (2) how to effectively learn distinguishable features despite single cross-entropy loss. We propose a novel Multi-modal Multi-layer Fusion Convolutional Neural Network (mmfCNN), which targets at finding a discriminative model for recognizing the subtle differences between live and spoof faces. The mmfCNN can fully use different information provided by diverse modalities, which is based on a weight-adaptation aggregation approach. Specifically, we utilize a multi-layer fusion model to further aggregate the features from different layers, which fuses the low-, mid- and high-level information from different modalities in a unified framework. Moreover, a novel Average Binary Center (ABC) loss is proposed to maximize the dissimilarity between the features of live and spoof faces, which helps to stabilize the training to generate a robust and discriminative model. Extensive experiments conducted on the CISIA-SURF and 3DMAD datasets verify the significance and generalization capability of the proposed method for the face anti-spoofing task. Code is available at: https://github.com/SkyKuang/Face-anti-spoofing.

Despite the exciting progress in image captioning, generating diverse captions for a given image remains as an open problem. Existing methods typically apply generative models such as Variational Auto-Encoder to diversify the captions, which however neglect two key factors of diverse expression, i.e., the lexical diversity and the syntactic diversity. To model these two inherent diversities in image captioning, we propose a Variational Structured Semantic Inferring model (termed VSSI-cap) executed in a novel structured encoder-inferer-decoder schema. VSSI-cap mainly innovates in a novel structure, i.e., Variational Multi-modal Inferring tree (termed VarMI-tree). In particular, conditioned on the visual-textual features from the encoder, the VarMI-tree models the lexical and syntactic diversities by inferring their latent variables (with variations) in an approximate posterior inference guided by a visual semantic prior. Then, a reconstruction loss and the posterior-prior KL-divergence are jointly estimated to optimize the VSSI-cap model. Finally, diverse captions are generated upon the visual features and the latent variables from this structured encoder-inferer-decoder model. Experiments on the benchmark dataset show that the proposed VSSI-cap achieves significant improvements over the state-of-the-arts.

Learning representations with diversified information remains as an open problem. Towards learning diversified representations, a new approach, termed Information Competing Process (ICP), is proposed in this paper. Aiming to enrich the information carried by feature representations, ICP separates a representation into two parts with different mutual information constraints. The separated parts are forced to accomplish the downstream task independently in a competitive environment which prevents the two parts from learning what each other learned for the downstream task. Such competing parts are then combined synergistically to complete the task. By fusing representation parts learned competitively under different conditions, ICP facilitates obtaining diversified representations which contain rich information. Experiments on image classification and image reconstruction tasks demonstrate the great potential of ICP to learn discriminative and disentangled representations in both supervised and self-supervised learning settings.

Hypergraph learning has been widely exploited in various image processing applications, due to its advantages in modeling the high-order information. Its efficacy highly depends on building an informative hypergraph structure to accurately and robustly formulate the underlying data correlation. However, the existing hypergraph learning methods are sensitive to non- Gaussian noise, which hurts the corresponding performance. In this paper, we present a noise-resistant hypergraph learning model, which provides superior robustness against various non- Gaussian noises. In particular, our model adopts low-rank representation to construct a hypergraph, which captures the globally linear data structure as well as preserving the grouping effect of highly-correlated data. We further introduce a correntropyinduced local metric to measure the reconstruction errors, which is particularly robust to non-Gaussian noises. Finally, the Frobenious-norm based regularization is proposed to combine with the low-rank regularizer, which enables our model to regularize the singular values of the coefficient matrix. By such, the non-zero coefficients are selected to generate a hyperedge set as well as the hyperedge weights. We have evaluated the proposed hypergraph model in the tasks of image clustering and semi-supervised image classification. Quantitatively, our scheme significantly enhances the performance of the state-of-the-art hypergraph models on several benchmark datasets.

Representation of videos is essential since it conveys an understanding of video content and enables many higher level tasks to be tackled efficiently. However, it is challenging to propose a rational representation for complex event videos, as most video information is either noisy or redundant. In this paper, we propose a compact event representation method that can concisely describe the inner modes of events. We deem that an optimal event representation scheme should reflect the long-term and high-level visual semantics (visual topics) of events, so different from previous frame-level video semantics representation methods and concept-based video representation methods, we investigate the problem from the perspective of segment-level video representations. We then present three appealing properties of segment-level visual semantics. Based on the observation, we propose different algorithms that rely on a novel deep-visual-word-based video encoding method to discover latent discriminative patterns of events. Finally, our multi-mode event representation is obtained by concatenating the discovered patterns as inner modes. We adopt our event representation for representative event parts mining, which can highlight the visual topics of events and remarkably prune the raw videos. We validate our event representation method based on complex event detection task. Experimental results on two standard benchmarking datasets, MED11 and CCV Dataset, show that the proposed method can significantly outperform the state-of-the-art approaches.

Most image captioning models focus on one-line (single image) captioning, where the correlations like relevance and diversity among group images (e.g., within the same album or event) are simply neglected, resulting in less accurate and diverse captions. Recent works mainly consider imposing the diversity during the online inference only, which neglect the correlation among visual structures in offline training. In this paper, we propose a novel group-based image captioning scheme (termed GroupCap), which jointly models the structured relevance and diversity among group images towards an optimal collaborative captioning. In particular, we first propose a visual tree parser (VP-Tree) to construct the structured semantic correlations within individual images. Then, the relevance and diversity among images are well modeled by exploiting the correlations among their tree structures. Finally, such correlations are modeled as constraints and sent into the LSTM-based captioning generator. We adopt an end-to-end formulation to train the visual tree parser, the structured relevance and diversity constraints, as well as the LSTM based captioning model jointly. To facilitate quantitative evaluation, we further release two group captioning datasets derived from the MS-COCO benchmark, serving as the first of their kind. Quantitative results show that the proposed GroupCap model outperforms the state-of-the-art and alternative approaches.

This work aims to estimate dense correspondences between the images from same visual class but with different geometries and visual similarities. This task is particularly challenging because (i) most image pairs have large intra-class variations, and (ii)their visual content is similar only on the high-level structure. To address these problems, this paper proposed a multilevel method to estimate per-pixel correspondences from high-level semantic to low-level structural details by the guidance of cycle-consistency. We utilize CNN feature pyramid to represent images level by level. Meanwhile, we introduce cycle-consistency to measure the reliability of flow vector, which further affects the guidance from higher level to lower level. The proposed method has been extensively evaluated on various challenging benchmarks. The results show that our method significantly outperforms the state-of-the-arts in terms of semantic flow accuracy.

Recognition of general actions has achieved great breakthroughs in recent years. However, in real-world applications, finer-grained action classification is often needed. The major challenge is that fine-grained actions usually share high similarities in both appearance and motion pattern, making it difficult to distinguish them with existing general action representation. To solve this problem, we introduce visual attention mechanism into the proposed descriptor, termed as Actionness-pooled Deep-convolutional Descriptor (ADD). Instead of pooling features uniformly from the entire video, we aggregate features in sub-regions that are more likely to contain actions according to actionness maps, which endow ADD with the capability of capturing the subtle differences between fine-grained actions. We conduct experiments on HIT Dances dataset, one of the few existing datasets for fine-grained action analysis. Quantitative results have demonstrated that ADD remarkably outperforms traditional two-stream representation. Extensive experiments on two general action benchmarks, JHMDB and UCF101, have additionally proved that combining ADD with end-to-end ConvNet can further boost the recognition performance.

Fine-grained object retrieval has attracted extensive research focus recently. Its state-of-the-art schemesare typically based upon convolutional neural network (CNN) features. Despite the extensive progress, two issues remain open. On one hand, the deep features are coarsely extracted at image level rather than precisely at object level, which are interrupted by background clutters. On the other hand, training CNN features with a standard triplet loss is time consuming and incapable to learn discriminative features. In this paper, we present a novel fine-grained object retrieval scheme that conquers these issues in a unified framework. Firstly, we introduce a novel centralized ranking loss (CRL), which achieves a very efficient (1,000times training speedup comparing to the triplet loss) and discriminative feature learning by a ?centralized? global pooling. Secondly, a weakly supervised attractive feature extraction is proposed, which segments object contours with top-down saliency. Consequently, the contours are integrated into the CNN response map to precisely extract features ?within? the target object. Interestingly, we have discovered that the combination of CRL and weakly supervised learning can reinforce each other. We evaluate the performance ofthe proposed scheme on widely-used benchmarks including CUB200-2011 and CARS196. We havereported significant gains over the state-of-the-art schemes, e.g., 5.4% over SCDA [Wei et al., 2017]on CARS196, and 3.7% on CUB200-2011.

3-D convolutional neural networks (3-D-convNets) have been very recently proposed for action recognition in videos, and promising results are achieved. However, existing 3-D-convNets has two “artificial” requirements that may reduce the quality of video analysis: 1) It requires a fixed-sized (e.g., 112 $\times$ 112) input video; and 2) most of the 3-D-convNets require a fixed-length input (i.e., video shots with fixed number of frames). To tackle these issues, we propose an end-to-end pipeline named Two-stream 3-D-convNet Fusion, which can recognize human actions in videos of arbitrary size and length using multiple features. Specifically, we decompose a video into spatial and temporal shots. By taking a sequence of shots as input, each stream is implemented using a spatial temporal pyramid pooling (STPP) convNet with a long short-term memory (LSTM) or CNN-E model, softmax scores of which are combined by a late fusion. We devise the STPP convNet to extract equal-dimensional descriptions for each variable-size shot, and we adopt the LSTM/CNN-E model to learn a global description for the input video using these time-varying descriptions. With these advantages, our method should improve all 3-D CNN-based video analysis methods. We empirically evaluate our method for action recognition in videos and the experimental results show that our method outperforms the state-of-the-art methods (both 2-D and 3-D based) on three standard benchmark datasets (UCF101, HMDB51 and ACT datasets).

In this paper, we present an automatic approach for on-line discovery of visual-lingual semantic fragments from weakly labeled Internet images. Instead of learning region-entity correspondences from well-labeled image-sentence pairs, our approach directly collects and enhances the weakly labeled visual contents from the Web and constructs an adaptive visual representation which automatically links generic lingual phrases to their related visual contents. To ensure reliable and efficient semantic discovery, we adopt non-parametric density estimation to re-rank the related visual instances and proposed a fast self-similarity-based quality assessment method to identify the high-quality semantic fragments. The discovered semantic fragments provide an adaptive joint representation for texts and images, based on which lingual-visual similarity can be defined for further co-analysis of heterogeneous multimedia data. Experimental results on semantic fragment quality assessment, sentence-based image retrieval, automatic multimedia insertion and ordering demonstrated the effectiveness of the proposed framework.The experiments show that the proposed methods can make effective use of the Web knowledge, and are able to generate competitive results compared to state-of-the-art approaches in various tasks.

In this paper, we proposed an integrated model of both semantic-aware and contrast-aware saliency (SCA) combining both bottom-up and top-down cues for effective eye fixation prediction. The proposed (SCA) model contains two pathways. The first pathway is a deep neural network customized for semantic-aware saliency, which aims to capture the semantic information in images, especially for the presence of meaningful objects and object parts. The second pathway is based on on-line feature learning and information maximization, which learns an adaptive representation for the input and discovers the high contrast salient patterns within the image context. The two pathways characterize both long-term and short-term attention cues and are integrated using maxima normalization. Experimental results on artificial images and several benchmark dataset demonstrate the superior performance and better plausibility of the proposed model over both classic approaches and recent deep models.

With the rapid rise of various e-commerce and social network platforms, users are generating large amounts of heterogeneous behavior data, such as purchasehistory, adding-to-favorite, adding-to-cart and click activities, and this kind of user behavior data is usually binary, only reflecting a user's action or inaction (i.e., implicit feedback data). Tensor factorization is a promising means of modeling heterogeneous user behaviors by distinguishing different behavior types. However, ambiguity arises in the interpretation of the unobserved user behavior records that mix both real negative examples and potential positive examples. Existing tensor factorization models either ignore unobserved examples or treat all of them as negative examples, leading to either poor prediction performance or huge computation cost. In addition, the distribution of positive examples w.r.t. behavior types is heavily skewed. Existing tensor factorization models would bias towards the type of behaviors with a large number of positive examples. In this paper, we propose a scalable probabilistic tensor factorization model (SPTF) for heterogeneous behavior data and develop a novel negative sampling technique to optimize SPTF by leveraging both observed and unobserved examples with much lower computational costs and higher modeling accuracy. To overcome the issue of the heavy skewness of the behavior data distribution, we propose a novel adaptive ranking-based positive sampling approach to speed up the model convergence and improve the prediction accuracy for sparse behavior types. Our proposed model optimization techniques enable SPTF to be scalable to large-scale behavior datasets. Extensive experiments have been conducted on a large-scale e-commerce dataset, and the experimental results show the superiority of our proposed SPTF model in terms of prediction accuracy and scalability.

Action Guidance (AG) aims at scoring how accurate the action is and giving guidance to the learners on how to correct their actions according to the standard instructive videos. AG has plenty of real-world applications such as sports training, rehabilitation treatment, and dance teaching. However, the problem of assessing the accuracy of action has almost no effective solution. In this paper, we describe a two-stage framework for action guidance. Firstly, we estimate the poses in the test video and standard video using person detection method and convolutional pose machines (CPMs). As for action guidance, we propose a network to compute the essential differences between two poses under different sizes, views, and locations. Extensive experiments with real-world and synthetic datasets demonstrate the effectiveness of our framework.

Context information plays an important role in object detection. DeepID concatenates the global context for classification, while Yolo, SSD, and Crafting use local context information for detection. In this paper, we propose a straightforward method to plug the global context information into the Faster RCNN Framework, namely the Skip Context Connection (SCC). We use SCC to inject the global context into the object representation which skips the RoiPooling layer rather than drops it. Therefore, it can not only leverage the context information but also keep the location accuracy from the RCNN framework. We proposed three principles to construct the SCC blocks: effectiveness means fewer parameters, additivity means the features possess the same meaning, and selectable means soft gated addition. We also evaluate several different SCC blocks. The Gated Additive SCC(GA-SCC) which satisfy the three principles get the best performance. Our experiment results on PASCAL VOC 2007 show that GA-SCC can get the steady 1% improvement over the traditional RCNN method.

Semantic information is important for video event detection. How to automatically discover, model, and utilize semantic information to facilitate video event detection has been a challenging problem. In this paper, we propose a novel hierarchical video event detection model, which deliberately unifies the processes of underlying semantics discovery and event modeling from video data. Specially, different from most of the approaches based on manually pre-defined concepts, we devise an effective model to automatically uncover video semantics by hierarchically capturing latent static-visual concepts in frame-level and latent activity concepts (i.e., temporal sequence relationships of static-visual concepts) in segment-level. The unified model not only enables a discriminative and descriptive representation for videos, but also alleviates error propagation problem from video representation to event modeling existing in previous methods. A max-margin framework is employed to learn the model. Extensive experiments on four challenging video event datasets, i.e., MED11, CCV, UQE50, and FCVID, have been conducted to demonstrate the effectiveness of the proposed method.

Dance is a unique and meaningful type of human expression, composed of abundant and various action elements. However, existing methods based on associated texts and spatial visual features have difficulty capturing the highly articulated motion patterns. To overcome this limitation, we propose to take advantage of the intrinsic motion information in dance videos to solve the video recommendation problem. We present a novel system that recommends dance videos based on a mid-level action representation, termed Dancelets. The Dancelets are used to bridge the semantic gap between video content and high-level concept, dance style, which plays a significant role in characterizing different types of dances. The proposed method executes automatic mining of dancelets with a concatenation of normalized cut clustering and linear discriminant analysis. This ensures that the discovered dancelets are both representative and discriminative. Additionally, to exploit the motion cues in videos, we employ motion boundaries as saliency priors to generate volumes of interest and extract C3D features to capture spatiotemporal information from the mid-level patches. Extensive experiments validated on our proposed large dance dataset, HIT Dances dataset, demonstrate the effectiveness of the proposed methods for dance style-based video recommendation.

Previous works on actor identification mainly focused on static features based on face identification and costume detection, without considering the abundant dynamic information contained in videos. In this paper, we propose a novel method to mine representative actions of each actor, and show the remarkable power of such actions for actor identification task. Videos are firstly divided into shots and represented by BoW based on spatial-temporal features. Then we integrate the prototype theory with SVM to rank the shots and obtain the representative actions. Our method for actor identification combines representative actions with actors' appearance. We validate the method on episodes of the TV series "The Big Bang Theory". The experimental results show that the representative actions are consistent with human judgements and can greatly improve the matching performance as complementary to existing handcrafted static features for actor identification.

Recently, neuron activations extracted from a pre-trained convolutional neural network (CNN) show promising performance in various visual tasks. However, due to the domain and task bias, using the features generated from the model pre-trained for image classification as image representations for instance retrieval is problematic. In this paper, we propose quartet-net learning to improve the discriminative power of CNN features for instance retrieval. The general idea is to map the features into a space where the image similarity can be better evaluated. Our network differs from the traditional Siamese-net in two ways. First, we adopt a double-margin contrastive loss with a dynamic margin tuning strategy to train the network which leads to more robust performance. Second, we introduce in the mimic learning regularization to improve the generalization ability of the network by preventing it from overfitting to the training data. Catering for the network learning, we collect a large-scale dataset, namely GeoPair, which consists of 68k matching image pairs and 63k non-matching pairs. Experiments on several standard instance retrieval datasets demonstrate the effectiveness of our method.

Most existing works on affective image classification tried to assign a dominant emotion category to an image. However, this is often insufficient, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the probability distribution of categorical image emotions. Firstly we extract commonly used features of different levels for each image. Then we formulize the emotion distribution prediction as a shared sparse leaning problem, which is optimized by iteratively reweighted least squares. Besides, we introduce three baseline algorithms. Experiments are carried out on a dataset of peer rated abstract paintings and the results demonstrate the superiority of our proposed method, as compared to some state-of-the-art approaches.

Recent developments in the field of image and video processing have led to a renewed interest in sketch correlated research. With that, there have emerged considerable solid evidence which revealed the significance of sketch to us. However, there have been few profound discussions on sketch based action analysis so far. In this paper, we present a framework of converting human actions to commendable sketches and discovering distinctive action sketches. The action sketches should satisfy three characteristics: sketchability, objective-ness and consistency. Primitive sketches are prepared according to the structured forests based fast edge detection. Meanwhile, we take advantage of Struck to accomplish adaptive object tracking in parallel. After that, we propose a method to ensure the spatio-temporal consistency between every sequential action sketches. On completion of previous stages, the process of distinctive mining of action sketches is carried out. The experimental results show that our approach has a promising potential.

Dance is a special and important type of action, composed of abundant and various action elements. However, the recommendation of dance videos on the web are still not well studied. It is hard to realize it in the way of traditional methods using associated texts or static features of video content. In this paper, we study the problem focusing on extraction and representation of action information in dances. We propose to recommend dance videos based on the automatically discovered ``Dance Styles'', which play a significant role in characterizing different types of dances. To bridge the semantic gap of video content and mid-level concept, style, we take advantage of a mid-level action representation method, and extract representative patches as ``Dancelets'', a sort of intermediation between videos and the concepts. Furthermore, we propose to employ Motion Boundaries as saliency priors and sparsely extract patches containing more representative information to generate a set of dancelet candidates. Dancelets are then discovered by Normalized-cut method, which is superior in grouping visually similar patterns into the same clusters. For the fast and effective recommendation, a random forest-based index is built, and the ranking results are derived according to the matching results in all the leaf notes. Extensive experiments validated on the web dance videos demonstrate the effectiveness of the proposed methods for dance style discovery and video recommendation based on styles.

In this study, the authors propose a collaborative composition model for automatically recommending suitable positions and poses in the scene of photography taken by amateurs. By analysing aesthetic‐aware features, the authors' strategy jointly takes attention and geometry composition into account to learn the aesthetic manifestation knowledge of professional photographers. Firstly, aesthetic composition representation exploits the strength of visual saliency to explicitly encode the spatial correlation of the professional photos. Secondly, ℓ2 regularised least square is adopted to constrain the representation coefficients, which provides a fast solution in selecting aesthetic candidates collaboratively. In addition, a novel confidence measure scheme is further designed based on reconstruction errors and the reference photos are updated adaptively according to the composition rules. Both qualitative and quantitative evaluations show that the model performs well for the portrait photographing recommendation.

In this paper, we develop a novel unsupervised crowd activity discovery algorithm aiming to automatically explore latent action patterns among crowd activities and partition them into meaningful clusters. Inspired by computational model of human vision system, we present a spatiotemporal saliency-based representation to simulate visual attention mechanism and encode human-focused components in an activity stream. Combining with feature pooling, we could obtain a more compact and robust activity representation. Based on the affinity matrix of activities, N-cut is performed to generate clusters with meaningful activity patterns. We carry out experiments on our proposed HIT-BJUT dataset and another public UMN dataset. The experimental results demonstrate that the proposed unsupervised discovery method is capable of automatically mining meaningful activities from large-scale video data with mixed crowd activities.

Recent progress on tracking has focused on designing robust statistical model or proposing effective appearance features to improve precision. This paper addresses another problem, namely the discovery and tracking of generic multi-object which have the similar appearance and motion pattern based on limited human annotations. We present a model-free tracking method that can automatically discover and track multi-object sharing the same spatial and motion structure, and update the structure during the tracking without prior acknowledge. The candidate objects are first selected by a SVM classifier trained on histogram-of-gradient (HOG) features. Then a segment algorithm is exploited to decide the suitable sizes of tracking boxes. The structure constrains are updated in a real-time manner according to the motion measure among the specified object and corresponding candidates. Experimental results reveal significant convenience and remarkable performance of our approach for the task of structure preserving multi-object discovery and tracking.

Covert attention is an mental ability to attend to a stimulus without shifting ones gaze towards it. The covert attention mechanism allows small spatial displacement during saccadic eye-movements, which we believed to be responsible for the existence of a large number of imperfectly allocated eye fixations in current saliency benchmark datasets. Inspired by this new finding, we propose to use spatial pooling to integrate cover attention into the current saliency models. We test our pooling-based boosting strategy for 20 state-of-the-art attention models on two well acknowledged fixation datasets (YORK-120 & MIT-1003). The experimental results show that our method can stably improve the performance of over 95% of the tested models in the eye fixation prediction task.

Emotions can be evoked in humans by images. Most previous works on image emotion analysis mainly used the elements-of-art-based low-level visual features. However, these features are vulnerable and not invariant to the different arrangements of elements. In this paper, we investigate the concept of principles-of-art and its influence on image emotions. Principles-of-art-based emotion features (PAEF) are extracted to classify and score image emotions for understanding the relationship between artistic principles and emotions. PAEF are the unified combination of representation features derived from different principles, including balance, emphasis, harmony, variety, gradation, and movement. Experiments on the International Affective Picture System (IAPS), a set of artistic photography and a set of peer rated abstract paintings, demonstrate the superiority of PAEF for affective image classification and regression (with about 5% improvement on classification accuracy and 0.2 decrease in mean squared error), as compared to the state-of-the-art approaches. We then utilize PAEF to analyze the emotions of master paintings, with promising results.

In this paper, we present a unified statistical framework for modeling both saccadic eye movements and visual saliency. By analyzing the statistical properties of human eye fixations on natural images, we found that human attention is sparsely distributed and usually deployed to locations with abundant structural information. This observations inspired us to model saccadic behavior and visual saliency based on super-Gaussian component (SGC) analysis. Our model sequentially obtains SGC using projection pursuit, and generates eye movements by selecting the location with maximum SGC response. Besides human saccadic behavior simulation, we also demonstrated our superior effectiveness and robustness over state-of-the-arts by carrying out dense experiments on synthetic patterns and human eye fixation benchmarks. Multiple key issues in saliency modeling research, such as individual differences, the effects of scale and blur, are explored in this paper. Based on extensive qualitative and quantitative experimental results, we show promising potentials of statistical approaches for human behavior research.

In this paper, we propose a computational model of visual representative ness by integrating cognitive theories of representative ness heuristics with computer vision and machine learning techniques. Unlike previous models that build their representative ness measure based on the visible data, our model takes the initial inputs as explicit positive reference and extend the measure by exploring the implic it negatives. Given a group of images that contains obvious visual concepts, we create a customized image ontology consisting of both positive and negative instances by mining the most related and confusable neighbors of the positive concept in ontological semantic knowledge bases. The representative ness of a new item is then determined by its likelihoods for both the positive and negative references. To ensure the effectiveness of probability inference as well as the cognitive plausibility, we discover the potential prototypes and treat them as an intermediate representation of semantic concepts. In the experiment, we evaluate the performance of representative ness models based on both human judgements and user-click logs of commercial image search engine. Experimental results on both Image Net and image sets of general concepts demonstrate the superior performance of our model against the state-of-the-arts.

This paper proposes a general method for size optimization in dense sampling to obtain a better representation of an image. Our method can be utilized to improve the performance of image classification and other tasks. We discuss the spatial consistency in global-scope restrained descriptors, by analyzing the appropriate sampling size. We apply the low rank method to solve the representative matrix of the descriptor sets at different scales, and obtain the optimized dense sampling size according to the lowest ranks of the representative matrices. Experimental results indicate that the proposed method gives an innovative and effective image representation, and it outperforms traditional dense sampling without size optimization.

In this paper, we present a unified statistical framework for modeling both saccadic eye movements and visual saliency. By analyzing the statistical properties of human eye fixations on natural images, we found that human attention is sparsely distributed and usually deployed to locations with abundant structural information. This new observations inspired us to model saccadic behavior and visual saliency based on Super Gaussian Component (SGC) analysis. The model sequentially obtains SGC using projection pursuit, and generates eye-movements by selecting the location with maximum SGC response. Beside human saccadic behavior simulation, we also demonstrated our superior effectiveness and robustness over state-of-the-arts by carrying out dense experiments on psychological patterns and human eye fixation benchmarks. These results also show promising potentials of statistical approaches for human behavior research.

In this paper, we present an intelligent portrait photographing framework for automatically recommending the suitable positions and poses in the scene of photography taken by amateurs. By analyzing aesthetic characteristics features, we propose a solution by constructing aesthetic composition representation which covers the attention composition and geometry composition to identify the underlying technique of professional photographer. First, we extract the attention composition feature of the professional photo by utilizing a visual saliency model. Then, a geometry composition feature is also presented to learn the spatial correlation. Finally, composition rules are applied to make appropriate pose and position. Experiments show our aesthetic composition representation performs well for portrait photographing recommendation.

Previous studies of sparse representation in multimedia research focus on developing reliable and efficient dictionary learning algorithms. Despite the sparse prior, how to integrate other related perceptual factors of human being into dictionary learning process was seldom studied. In this paper, we investigate the influence of image memorability for human-centralized sparse representation. Based on the results of a photo memory game, we are able to quantitatively characterize an image's memorability which allows us to train sparse bases from the most memorable images instead of randomly selected natural images. We believed that such kind of basis is more consistent with neural networks in human brain and hence can better predict where human looks. To test our hypothesis, we choose human eye-fixation prediction problem for quantitative evaluation. The experimental results demonstrate the superior performance of our Memorable Basis compared to traditional sparse basis trained from unselected images.

With the great popularity and extensive application of Kinect, the Internet is sharing more and more depth data. To effectively use plenty of depth data would make great sense. In this paper, we propose a generalized dynamic depth data matching framework for action retrieval. Firstly we focus on single depth image matching utilizing both depth and shape feature. The depth feature used in our method is straightforward but proved to be very effective and robust for distinguishing various human actions. Then, we adopt shape context, which is widely used in shape matching, in order to strengthen the robustness of our matching strategy. Finally, we utilize Dynamic Time Warping to measure temporal similarity between two depth video sequences. Experiments based on a dataset of 17 classes of actions from 10 different individuals demonstrate the effectiveness and robustness of our proposed matching strategy.

A visual codebook serves as a fundamental component in many state-of-the-art computer vision systems. Most existing codebooks are built based on quantizing local feature descriptors extracted from training images. Subsequently, each image is represented as a high-dimensional bag-of-words histogram. Such highly redundant image description lacks efficiency in both storage and retrieval, in which only a few bins are nonzero and distributed sparsely. Furthermore, most existing codebooks are built based solely on the visual statistics of local descriptors, without considering the supervise labels coming from the subsequent recognition or classification tasks. In this paper, we propose a task-dependent codebook compression framework to handle the above two problems. First, we propose to learn a compression function to map an originally high-dimensional codebook into a compact codebook while maintaining its visual discriminability. This is achieved by a codeword sparse coding scheme with Lasso regression, which minimizes the descriptor distortions of training images after codebook compression. Second, we propose to adapt our codebook compression to the subsequent recognition or classification tasks. This is achieved by introducing a label constraint kernel (LCK) into our compression loss function. In particular, our LCK can model heterogeneous kinds of supervision, i.e., (partial) category labels, correlative semantic annotations, and image query logs. We validated our codebook compression in three computer vision tasks: 1) object recognition in PASCAL Visual Object Class 07; 2) near-duplicate image retrieval in UKBench; and 3) web image search in a collection of 0.5 million Flickr photographs. Our compressed codebook has shown superior performances over several state-of-the-art supervised and unsupervised codebooks.

How can interest point detectors benefit from contextual cues? In this articles, we introduce a context-aware semi-local detector (CASL) framework to give a systematic answer with three contributions: (1) We integrate the context of interest points to recurrently refine their detections. (2) This integration boosts interest point detectors from the traditionally local scale to a semi-local scale to discover more discriminative salient regions. (3) Such context-aware structure further enables us to bring forward category learning (usually in the subsequent recognition phase) into interest point detection to locate category-aware, meaningful salient regions. Our CASL detector consists of two phases. The first phase accumulates multiscale spatial correlations of local features into a difference of contextual Gaussians (DoCG) field. DoCG quantizes detector context to highlight contextually salient regions at a semi-local scale, which also reveals visual attentions to a certain extent. The second phase locates contextual peaks by mean shift search over the DoCG field, which subsequently integrates contextual cues into feature description. This phase enables us to integrate category learning into mean shift search kernels. This learning-based CASL mechanism produces more category-aware features, which substantially benefits the subsequent visual categorization process. We conducted experiments in image search, object characterization, and feature detector repeatability evaluations, which reported superior discriminability and comparable repeatability to state-of-the-art works.

Saliency detection is essential to visual attention modelling and various computer vision tasks. Representation and measurement are two important issues for saliency models. Good representation and reasonable measurement are both critical issues in modelling visual saliency mechanism. For every input image, we obtain a self-adaptive dictionary that describes the image content effectively and image prior that forces sparsity in every location in the image using the K-SVD algorithm. For saliency measurement, background firing rate (BFR) is defined for each sparse features and it is followed by feature activation rate (FAR) computation to measure the bottom-up visual saliency.

In this paper, we propose a novel affective video classification method based on facial expression recognition by learning the spatio-temporal feature fusion of actors' and viewers' facial expressions. For spatial features, we integrate Haar-like features into compositional ones according to the features' correlation, and train a mid classifier during the period. Then this process is embedded into improved AdaBoost learning algorithm to obtain spatial features. And for temporal feature fusion, we adopt hidden dynamic conditional random fields (HDCRFs) based on HCRFs by introducing time dimension variable. Finally spatial features are embedded into HDCRFs to recognize facial expressions. Experiments on the well-known Cohn-Kanada database show that the proposed method has a promising recognition performance. And affective classification experimental results on our own videos show that most subjects are satisfied with the classification results.

Modern clothes are designed based on various visual elements of different fashion styles. Traditional vision-based clothes recommendation methods focused on searching clothes which are similar with user preferred samples in the aspects of colors and partial shape elements. In this paper, we propose a method of recommending clothes by mining visual elements of different fashion styles. Independent Component Analysis (ICA) is employed to extract sparse features, and then Term-Frequency (TF) analysis is applied to discover visual elements from these independent components. Finally, we test three ranking metrics for clothes recommendation including Euclidian distance of TFs, Cosine distance of TFs and Minimum TF. Experimental results based on web commercial images demonstrate the effectiveness of the proposed method.

Web videos such as YouTube are hard to obtain sufficient precisely labeled training data and analyze due to the complex ontology. To deal with these problems, we present a hierarchical web video classification framework by learning heterogeneous web data, and construct a bottom-up semantic forest of video concepts by learning from meta-data. The main contributions are two-folds: firstly, analysis about middle-level concepts' distribution is taken based on data collected from web communities, and a concepts redistribution assumption is made to build effective transfer learning algorithm. Furthermore, an AdaBoost-Like transfer learning algorithm is proposed to transfer the knowledge learned from Flickr images to YouTube video domain and thus it facilitates video classification. Secondly, a group of hierarchical taxonomies named Semantic Forest are mined from YouTube and Flickr tags which reflect better user intention on the semantic level. A bottom-up semantic integration is also constructed with the help of semantic forest, in order to analyze video content hierarchically in a novel perspective. A group of experiments are performed on the dataset collected from Flickr and YouTube. Compared with state-of-the-arts, the proposed framework is more robust and tolerant to web noise.

In this paper, we proposed a fast and robust unsupervised framework for anomaly detection and localization in crowed scenes. Our method avoids modeling the normal state of the crowds which is a very complex task due to the large within class variance of the normal target appearance and motion patterns. For each video frame, we extract the spatial temporal features of 3D blocks and generate the saliency map using a block-based center-surround difference operator. Then, motion vector matrix is obtained by adaptive rood pattern search block-matching algorithm and distance normalization. Attractive motion disorder descriptor is proposed to measure the global intensity of anomalies in the scene. Finally, we classify the frames into normal and anomalous ones by a binary classifier. In the experiments, we compared our method against several state-of-the-art approaches on UCSD dataset which is a widely used anomaly detection and localization benchmark. As the only unsupervised approach, our method outputs competitive results with near real-time processing speed

Most previous works on video indexing and recommendation were only based on the content of video itself, without considering the affective analysis of viewers, which is an efficient and important way to reflect viewers' attitudes, feelings and evaluations of videos. In this paper, we propose a novel method to index and recommend videos based on affective analysis, mainly on facial expression recognition of viewers. We first build a facial expression recognition classifier by embedding the process of building compositional Haar-like features into hidden conditional random fields (HCRFs). Then we extract viewers' facial expressions frame by frame through the videos, collected from the camera when viewers are watching videos, to obtain the affections of viewers. Finally, we draw the affective curve which tells the process of affection changes. Through the curve, we segment each video into affective sections, give the indexing result of the videos, and list recommendation points from views' aspect. Experiments on our collected database from the web show that the proposed method has a promising performance.

Visual vocabulary serves as a fundamental component in many computer vision tasks, such as object recognition, visual search, and scene modeling. While state-of-the-art approaches build visual vocabulary based solely on visual statistics of local image patches, the correlative image labels are left unexploited in generating visual words. In this work, we present a semantic embedding framework to integrate semantic information from Flickr labels for supervised vocabulary construction. Our main contribution is a Hidden Markov Random Field modeling to supervise feature space quantization, with specialized considerations to label correlations: Local visual features are modeled as an Observed Field, which follows visual metrics to partition feature space. Semantic labels are modeled as a Hidden Field, which imposes generative supervision to the Observed Field with WordNet-based correlation constraints as Gibbs distribution. By simplifying the Markov property in the Hidden Field, both unsupervised and supervised (label independent) vocabularies can be derived from our framework. We validate our performances in two challenging computer vision tasks with comparisons to state-of-the-arts: (1) Large-scale image search on a Flickr 60,000 database; (2) Object recognition on the PASCAL VOC database.

Mining actor correlations from TV series enables semantic level video understanding and facilitates users to conduct correlation-based query. In this paper, we introduce a graph-based actor correlations mining framework, which serves as the first attempt for effective actor association presentation and concurrence search. We leverage face detection and tracking to locate actors with 2D-PCA detector as pretreatment. To measure the actor association into a unified graph, we propose a context-based actor correlations hierarchical parsing approach, which considers video structure and hierarchical concurrence to refine actor association in our graph modeling. We not only can accomplish actor correlations mining, but also can acquire higher semantic information according to concurrence change. We present the actor correlation mining results in a graph-based interface to enable efficient users' navigation and search.

Human vision system acquires essential information from the environment by sequentially sampling visual contents at important locations under the control of selective attention mechanism. We propose that bottom-up saliency is not based on global statistics but on information sampled at prior eye fixations. Our model calculates visual saliency using sequential eye fixation probability. However, the proposed model needs fixation priors, which are hard to simulate given current fixation data and experimental conditions. An approximation is proposed to generate a single saliency map by fusing all possible conditions of fixation prior. Our method outperforms all state-of-the-art models in predicting eye fixations, and shows reasonable response to various psychological patterns.

Representation and measurement are two important issues for saliency models. Different with previous works that learnt sparse features from large scale natural statistics, we propose to learn features from short-term statistics of single images. For saliency measurement, we define background firing rate (BFR) for each sparse feature, and then we propose to use feature activation rate (FAR) to measure the bottom-up visual saliency. The proposed FAR measure is biological plausible and easy to compute, also with satisfied performance. Experiments on human eye fixations and psychological patterns demonstrate the effectiveness and robustness of our proposed method.

This paper presents a novel texture description approach, which is robust to variances in rotation, scale and illumination in images, to classify the texture of images. A limitation with traditional methods is that they are more or less sensitive to the mentioned changes in images. To overcome this problem, we propose a novel Local Haar Binary Pattern (LHBP) based framework to ensure invariance in global rotation, scale, and light change. Our method consists of two components: feature extraction and scale self-adaptive classification. The global rotation invariant LHBP histogram features are extracted against the variances of illumination and global rotation, and the scale self-adaptive strategy is used for optimizing the classification of different scale textures. Evaluation results on Outex and Brodatz databases illustrate the significant advantages of the proposed approach over existing algorithms.

Does there exist a compact set of keywords that can completely and effectively cover the image annotation problem by expanding from it? In this paper, we answer this question by presenting a complete set framework for image annotation, which is motivated by the existence of semantic ontology. To generate this set, we propose a cross model optimization strategy from both textual and visual information for topic decomposition, based on a so-called Bipartite LSA model, which minimize multimodal error energy functions in a probabilistic Latent Semantic Analysis model. To achieve complete set based annotation, we present a Gaussian-Kernel-Generative process based keyword generation procedure, which analogizes keyword annotation in a probabilistic generative manner. A group of experiments is performed on Washington University image database and 80,000 Flickr images with comparisons to the state-of-the-arts. Finally, potential advantages and future improvements of our framework are discussed outside the scope of topic modeling.

It is difficult to be satisfied for automatic photo assessment using only low level visual features such as brightness, lighting, hue, contrast, color distribution and so on. Instead of using low level visual features, we present a novel computational visual attention model to assess photos. Firstly, a face-sensitive saliency map analysis is deployed to estimate attention distribution. Then, a Rate of Focused Attention (RFA) measurement is proposed to quantify photo quality. By integrating top-down supervision into the visual attention model, we further achieve personalized photo assessment to take user preference into quality evaluation, which can be extended into object or semantic oriented photo assessment scenarios. Experiments on personal photo albums with comparison to ground-truth user evaluations demonstrate the effeteness of the proposed method.

Two main restrictions exist in state-of-the-art text detection algorithms: 1. Illumination variance; 2. Text-background contrast variance. This paper presents a robust text characterization approach based on local Haar binary pattern (LHBP) to address these problems. Based on LHBP, a coarse-to-fine detection framework is presented to precisely locate text lines in scene images. Firstly, threshold-restricted local binary pattern is extracted from high-frequency coefficients of pyramid Haar wavelet. It preserves and uniforms inconsistent text-background contrasts while filtering gradual illumination variations. Subsequently, we propose a directional correlation analysis (DCA) approach to filter non-directional LHBP regions for locating candidate text regions. Finally, using LHBP histogram, an SVM-based post-classification is presented to refine detection results. Experimental results on ICDAR 03 demonstrate the effectiveness and robustness of our proposed method.

From visual perception viewpoint, actions in videos can capture high-level semantics for video content understanding and retrieval. However, action-level video retrieval meets great challenges, due to the interferences from global motions or concurrent actions, and the difficulties in robust action describing and matching. This paper presents a content-based action retrieval framework to enable effective search of near-duplicated actions in large-scale video database. Firstly, we present an attention shift model to distill and partition human-concerned saliency actions from global motions and concurrent actions. Secondly, to characterize each saliency action, we extract 3D-SIFT descriptor within its spatial-temporal region, which is robust against rotation, scale, and view point variances. Finally, action similarity is measured using Dynamic Time Warping (DTW) distance to offer tolerance for action duration variance and partial motion missing. Search efficiency in large-scale dataset is achieved by hierarchical descriptor indexing and approximate nearest-neighbor search. In validation, we present a prototype system VILAR to facilitate action search within "Friends" soap operas with excellent accuracy, efficiency, and human perception revealing ability.