Face videos accompanied by audio have become integral to our daily lives, while they often suffer from complex degradations. Most face video restoration methods neglect the intrinsic correlations between visual and audio features, particularly in the mouth region. Several audio-aided face video restoration methods have been proposed, but they only focus on compression artifact removal. In this paper, we propose a General Audio-assisted face Video restoration Network (GAVN) to address various types of streaming video distortions via identity and temporal complementary learning. Specifically, GAVN first captures inter-frame temporal features in the low-resolution space to restore frames coarsely and save computational cost. Then, GAVN extracts intra-frame identity features in the high-resolution space with the assistance of audio signals and face landmarks to restore more facial details. Finally, the reconstruction module integrates temporal features and identity features to generate high-quality face videos. Experimental results demonstrate that GAVN outperforms the existing state-of-the-art methods on face video compression artifact removal, deblurring, and super-resolution.

The data scaling law has significantly enhanced large multi-modal models (LMMs) performance across various downstream tasks. However, in the domain of perceptual video quality assessment (VQA), the potential of data scaling remains unprecedented due to the scarcity of labeled resources and the insufficient scale of datasets. To address this, we propose OmniVQA, a framework designed to efficiently build high-quality, machine-dominated synthetic multi-modal instruction databases (MIDBs) for VQA. We then scale up to create OmniVQA-Chat-400K, the largest dataset in the VQA field concurrently. Our focus is on the technical and aesthetic quality dimensions, with abundant in-context instruction data to provide fine-grained VQA knowledge. Additionally, we build the OmniVQA-MOS-20K dataset to enhance the model's quantitative quality rating capabilities. We then introduce a complementary training strategy that effectively leverages the knowledge from datasets for different tasks. Furthermore, we propose the OmniVQA-FG (fine-grain)-Benchmark to evaluate the fine-grained performance of models. Our results demonstrate that our models achieve state-of-the-art performance in both tasks.

High-fidelity 3D meshes can be tokenized into one-dimension (1D) sequences and directly modeled using autoregressive approaches for faces and vertices. However, existing methods suffer from insufficient resource utilization, resulting in slow inference and the ability to handle only small-scale sequences, which severely constrains the expressible structural details. We introduce the Latent Autoregressive Network (LANE), which incorporates compact autoregressive dependencies in the generation process, achieving a 6× improvement in maximum generatable sequence length compared to existing methods. To further accelerate inference, we propose the Adaptive Computation Graph Reconfiguration (AdaGraph) strategy, which effectively overcomes the efficiency bottleneck of traditional serial inference through spatiotemporal decoupling in the generation process. Experimental validation demonstrates that LANE achieves superior performance across generation speed, structural detail, and geometric consistency, providing an effective solution for high-quality 3D mesh generation.

Large multimodal models (LMMs) have demonstrated remarkable capabilities across a wide range of tasks, however their knowledge and abilities in the cross-view geo-localization and pose estimation domains remain unexplored, despite potential benefits for navigation, autonomous driving, outdoor robotics, etc. To bridge this gap, we introduce GeoX-Bench, a comprehensive Benchmark designed to explore and evaluate the capabilities of LMMs in cross-view Geo-localization and pose estimation. Specifically, GeoX-Bench contains 10,859 panoramic-satellite image pairs spanning 128 cities in 49 countries, along with corresponding 755,976 question-answering (QA) pairs. Among these, 42,900 QA pairs are designated for benchmarking, while the remaining are intended to enhance the capabilities of LMMs. Based on GeoX-Bench, we evaluate the capabilities of 25 state-of-the-art LMMs on cross-view geo-localization and pose estimation tasks, and further explore the empowered capabilities of instruction-tuning. Our benchmark demonstrate that while current LMMs achieve impressive performance in geo-localization tasks, their effectiveness declines significantly on the more complex pose estimation tasks, highlighting a critical area for future improvement, and instruction-tuning LMMs on the training data of GeoX-Bench can significantly improve the cross-view geo-sense abilities.

Pairwise learning underpins implicit collaborative filtering, yet its effectiveness is often hindered by sparse supervision, noisy interactions, and popularity-driven exposure bias. In this paper, we propose Variational Bayesian Personalized Ranking (VarBPR), a tractable variational framework for implicit-feedback pairwise learning that offers principled exposure controllability and theoretical interpretability. VarBPR reformulates pairwise learning as variational inference over discrete latent indexing variables, explicitly modeling noise and indexing uncertainty, and divides training into two stages: variational inference, which solve variational posteriors, and variational learning, which updates model parameters based on these posteriors. In the variational inference stage, we develop a variational formulation that integrates preference alignment, denoising, and popularity debiasing under a unified ELBO/regularization objective, deriving closed-form posteriors with clear control semantics: the prior encodes a target exposure pattern, while temperature/regularization strength controls posterior-prior adherence. As a result, exposure controllability becomes an endogenous and interpretable outcome of variational inference. In the variational learning stage, we propose a posterior-compression objective that reduces the ideal ELBO's computational complexity from polynomial to linear, with the approximation justified by an explicit Jensen-gap upper bound. Theoretically, we provide interpretable generalization guarantees by identifying a structural error component and revealing the opportunity cost of prioritizing certain exposure patterns (e.g., long-tail), offering a concrete analytical lens for designing controllable recommender systems. Empirically, We validate VarBPR across popular backbones; it demonstrates consistent gains in ranking accuracy, enables controlled long-tail exposure, and preserves the linear-time complexity of BPR.

The recent rapid development of video generation technology has led to a significant demand for quality assessment of the latest AI-generated videos. However, current supervised approaches depend on expensive and quickly outdated human scores, and label-free methods overlook the general distortions of AI-generated videos. To address these limitations, we introduce LMVQ, a Label-free Metric-learning framework for general AI-generated Video Quality assessment of three dimensions, spatial, temporal, and alignment. The LMVQ is the first to introduce sample degradations specially designed for AIGC-specific distortions, and constructs a comprehensive training set through two complementary sample generation strategies. It then employs two synergistic modules, the Intra-Quality Token Transformer (IQ-Trans), which explicitly refines dimension-specific quality representations, and the Inter-Quality Mixture of Experts (IQ-MoE), which fuses interactions across multiple quality dimensions. Finally, a Multi-Proxy Metric-Learning (MPML) strategy aligns the learned representations with multi-dimensional quality scores and constrains the model to learn discriminative quality-aware representations. Extensive experiments on four public AIGC-VQA benchmarks show that MPML outperforms previous label-free methods by over 20%, and greatly narrows the gap with supervised methods. This provides a scalable, adaptive foundation for evaluating the ever-evolving quality of AI-generated videos.

The rapid advancement of AI generation technologies has led to the widespread use of AI-generated multimedia content, including images, videos, and 3D contents, across various applications. While significant progress has been made in quality evaluation for 2D content, evaluating the quality of 3D content synthesized from single image remains an underexplored problem. To bridge this gap, we introduce the first comprehensive subjective evaluation database tailored for assessing the quality of 3D content generated from single image. Our database, named AIGC-SI23DCQA, includes three distinct categories of input images, i.e., realistic images, AI-generated images, and computer graphic (CG) images, with 100 images in each category. Using five representative single-image-to-3D algorithms, we produce 1,500 3D contents and collect 94,500 annotations across three quality dimensions, including texture fidelity, shape accuracy, and overall quality. Based on the constructed database, we first benchmark and evaluate the performance of existing quality assessment methods revealing their limitations in addressing this novel task. Thus, we further propose a novel objective quality assessment method, termed I3DQA, for effective single-image-to-3D content quality assessment. Specifically, I3DQA first extracts the reference features from the source image, and the multi-modal features from the generated 3D content, including the projected video, patches, and large-multimodal model (LMM) features. These features are integrated through symmetric transformer blocks, enabling effective quality-related feature fusion and score prediction. Extensive experiments demonstrate the superior performance of our method and validate the effectiveness of its components. This work provides a foundational resource and a robust framework for advancing research in this emerging field, and our database and model are released at https://github.com/ZedFu/SI23DCQA.

Scribble supervision reduces annotation costs in multi-organ segmentation. However, its sparsity results in insufficient supervision for most regions and inadequate feature learning in hard areas (e.g., organ boundaries). These hard areas cause model confirmation bias and high epistemic uncertainty, which existing methods fail to address. To overcome these core challenges, we propose an epistemic-driven hardness-adaptive focusing framework. This framework establishes a self-improving loop: quantified epistemic uncertainty guides hard sample generation, while hard sample learning and feature alignment jointly reduce epistemic uncertainty. Specifically, we first propose a phase-adaptive hardness-aware loss function to quantify epistemic uncertainty and generate dynamic hardness maps during training. Based on these maps, we employ a distribution-divergence-aware copy-paste operation to create hard samples, which are progressively incorporated into learning to reduce epistemic uncertainty. Furthermore, we introduce feature distribution alignment to mitigate bias and epistemic uncertainty by aligning organ-specific hard regions with global features. Extensive experiments on multi-organ CT and ultrasound datasets demonstrate the competitiveness and effectiveness of our method. The framework's generalizability and robustness are further validated under cross-dataset and noise-corrupted scenarios. This work offers a practical solution for clinical applications where annotation efficiency is critical.

Zicheng Zhang, Xiangyu Zhao, Xinyu Fang, Chunyi Li, Xiaohong Liu, Xiongkuo Min, Haodong Duan, Kai Chen, Guangtao Zhai. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Deepfake detection is a long-established research topic vital for mitigating the spread of malicious misinformation. Unlike prior methods that provide either binary classification results or textual explanations separately, we introduce a novel method capable of generating both simultaneously. Our method harnesses the multi-modal learning capability of the pre-trained CLIP and the unprecedented interpretability of large language models (LLMs) to enhance both the generalization and explainability of deep-fake detection. Specifically, we introduce a multi-modal face forgery detector (M2F2-Det) that employs tailored face forgery prompt learning, incorporating the pre-trained CLIP to improve generalization to unseen forgeries. Also, M2F2-Det incorporates an LLM to provide detailed textual explanations of its detection decisions, enhancing interpretability by bridging the gap between natural language and subtle cues of facial forgeries. Empirically, we evaluate M2F2-Det on both detection and explanation generation tasks, where it achieves state-of-the-art performance, demonstrating its effectiveness in identifying and explaining diverse forgeries. Source code is available at $\color{magenta}{link}$.

Existing salient object detection (SOD) models primarily resort to convolutional neural networks (CNNs) and Transformers. However, the limited receptive fields of CNNs and quadratic computational complexity of transformers both constrain the performance of current models on discovering attention-grabbing objects. The emerging state space model, namely Mamba, has demonstrated its potential to balance global receptive fields and computational complexity. Therefore, we propose a novel unified framework based on the pure Mamba architecture, dubbed saliency Mamba (Samba), to flexibly handle general SOD tasks, including RGB/RGB-D/RGB-T SOD, video SOD (VSOD), and RGB-D VSOD. Specifically, we rethink Mamba’s scanning strategy from the perspective of SOD, and identify the importance of maintaining spatial continuity of salient patches within scanning sequences. Based on this, we propose a saliency-guided Mamba block (SGMB), incorporating a spatial neighboring scanning (SNS) algorithm to preserve spatial continuity of salient patches. Additionally, we propose a context-aware upsampling (CAU) method to promote hierarchical feature alignment and aggregations by modeling contextual dependencies Experimental results show that our Samba outperforms existing methods across five SOD tasks on 21 datasets with lower computational cost, confirming the superiority of introducing Mamba to the SOD areas. Our code is available at https://github.com/Jia-Hao999/Samba.

Multi-object images are prevalent in various real-world scenarios, including augmented reality, advertisement design, and medical imaging. Efficient and precise editing of these images is critical for these applications. With the advent of Stable Diffusion (SD), high-quality image generation and editing have entered a new era. However, existing methods often struggle to consider each object both individually and part of the whole image editing, both of which are crucial for ensuring consistent quantity perception, resulting in suboptimal perceptual performance. To address these challenges, we propose MoEdit, an auxiliaryfree multi-object image editing framework. MoEdit facilitates high-quality multi-object image editing in terms of style transfer, object reinvention, and background regeneration, while ensuring consistent quantity perception between inputs and outputs, even with a large number of objects. To achieve this, we introduce the Feature Compensation (FeCom) module, which ensures the distinction and separability of each object attribute by minimizing the in-between interlacing. Additionally, we present the Quantity Attention (QTTN) module, which perceives and preserves quantity consistency by effective control in editing, without relying on auxiliary tools. By leveraging the SD model, MoEdit enables customized preservation and modification of specific concepts in inputs with high quality. Experimental results demonstrate that our MoEdit achieves State-Of-The-Art (SOTA) performance in multi-object image editing. Data and codes are available at https://github.com/Tear-kitty/MoEdit.

Image Quality Assessment (IQA) based on human subjective preferences has undergone extensive research in the past decades. However, with the development of communication protocols, the visual data consumption volume of machines has gradually surpassed that of humans. For machines, the preference depends on downstream tasks such as segmentation and detection, rather than visual appeal. Considering the huge gap between human and machine visual systems, this paper proposes the topic: Image Quality Assessment for Machine Vision for the first time. Specifically, we (1) defined the subjective preferences of machines, including downstream tasks, test models, and evaluation metrics; (2) established the Machine Preference Database (MPD), which contains 2.25M fine-grained annotations and 30k reference/distorted image pair instances; (3) verified the performance of mainstream IQA algorithms on MPD. Experiments show that current IQA metrics are human-centric and cannot accurately characterize machine preferences. We sincerely hope that MPD can promote the evolution of IQA from human to machine preferences. Project page is on: https://github.com/lcysyzxdxc/MPD.

Traditional spatiotemporal models generally rely on task-specific architectures, which limit their generalizability and scalability across diverse tasks due to domain-specific design requirements. In this paper, we introduce UniSTD, a unified Transformer-based framework for spatiotemporal modeling, which is inspired by advances in recent foundation models with the two-stage pretraining-then-adaption paradigm. Specifically, our work demonstrates that task-agnostic pretraining on 2D vision and vision-text datasets can build a generalizable model foundation for spatiotemporal learning, followed by specialized joint training on spatiotemporal datasets to enhance task-specific adaptability. To improve the learning capabilities across domains, our framework employs a rank-adaptive mixture-of-expert adaptation by using fractional interpolation to relax the discrete variables so that can be optimized in the continuous space. Additionally, we introduce a temporal module to incorporate temporal dynamics explicitly. We evaluate our approach on a large-scale dataset covering 10 tasks across 4 disciplines, demonstrating that a unified spatiotemporal model can achieve scalable, cross-task learning and support up to 10 tasks simultaneously within one model while reducing training costs in multi-domain applications. Code will be available at https://github.com/1hunters/UniSTD.

Existing real-world image dehazing methods primarily attempt to fine-tune pre-trained models or adapt their inference procedures, thus heavily relying on the pre-trained models and associated training data. Moreover, restoring heavily distorted information under dense haze requires generative diffusion models, whose potential in de-hazing remains underutilized partly due to their lengthy sampling processes. To address these limitations, we introduce a novel hazing-dehazing pipeline consisting of a Realistic Hazy Image Generation framework (HazeGen) and a Diffusion-based Dehazing framework (DiffDehaze). Specifically, HazeGen harnesses robust generative diffusion priors of real-world hazy images embedded in a pre-trained text-to-image diffusion model. By employing specialized hybrid training and blended sampling strategies, HazeGen produces realistic and diverse hazy images as high-quality training data for DiffDehaze. To alleviate the inefficiency and fidelity concerns associated with diffusion-based methods, DiffDehaze adopts an Accelerated Fidelity-Preserving Sampling process (AccSamp). The core of AccSamp is the Tiled Statistical Alignment Operation (AlignOp), which can provide a clean and faithful dehazing estimate within a small fraction of sampling steps to reduce complexity and enable effective fidelity guidance. Extensive experiments demonstrate the superior dehazing performance and visual quality of our approach over existing methods. The code is available at https://github.com/ruiyi-w/Learning-Hazing-to-Dehazing.

With the rising interest in research on Large Multi-modal Models (LMMs) for video understanding, many studies have emphasized general video comprehension capabilities, neglecting the systematic exploration into video quality understanding. To address this oversight, we introduce Q-Bench-Video in this paper, a new benchmark specifically designed to evaluate LMMs' proficiency in discerning video quality. a) To ensure video source diversity, Q-Bench-Video encompasses videos from natural scenes, AI-generated content (AIGC), and computer graphics (CG). b) Building on the traditional multiple-choice questions format with the Yes-or-No and What-How categories, we include Open-ended questions to better evaluate complex scenarios. Additionally, we incorporate the video pair quality comparison question to enhance comprehensiveness. c) Beyond the traditional Technical, Aesthetic, and Temporal distortions, we have expanded our evaluation aspects to include the dimension of AIGC distortions, which addresses the increasing demand for video generation. Finally, we collect a total of 2,378 question-answer pairs and test them on 12 open-source & 5 proprietary LMMs. Our findings indicate that while LMMs have a foundational understanding of perceptual video quality, their performance remains incomplete and imprecise, with a notable discrepancy compared to the performance of human beings. Through Q-Bench-Video, we seek to catalyze community interest, stimulate further research, and unlock the untapped potential of LMMs to close the gap in video quality understanding.

Evaluating text-to-vision content hinges on two crucial aspects: visual quality and alignment. While significant progress has been made in developing objective models to assess these dimensions, the performance of such models heavily relies on the scale and quality of human annotations. According to Scaling Law, increasing the number of human-labeled instances follows a predictable pattern that enhances the performance of evaluation models. Therefore, we introduce a comprehensive dataset designed to Evaluate Visual quality and Alignment Level for text-to-vision content (Q-EVAL-100K), featuring the largest collection of human-labeled Mean Opinion Scores (MOS) for the mentioned two aspects. The Q-EVAL-100K dataset encompasses both text-to-image and text-to-video models, with 960K human annotations specifically focused on visual quality and alignment for 100K instances (60K images and 40K videos). Leveraging this dataset with context prompt, we propose Q-Eval-Score, a unified model capable of evaluating both visual quality and alignment with special improvements for handling long-text prompt alignment. Experimental results indicate that the proposed Q-Eval-Score achieves superior performance on both visual quality and alignment, with strong generalization capabilities across other benchmarks. These findings highlight the significant value of the Q-EVAL-100K dataset. Data and codes will be available at https://github.com/zzc-1998/Q-Eval.

The rapid development of Multi-modality Large Language Models (MLLMs) has significantly influenced various aspects of industry and daily life, showcasing impressive capabilities in visual perception and understanding. However, these models also exhibit hallucinations, which limit their reliability as AI systems, especially in tasks involving low-level visual perception and understanding. We believe that hallucinations stem from a lack of explicit self-awareness in these models, which directly impacts their overall performance. In this paper, we aim to define and evaluate the self-awareness of MLLMs in low-level visual perception and understanding tasks. To this end, we present QL-Bench, a benchmark settings to simulate human responses to low-level vision, investigating self-awareness in low-level visual perception through visual question answering related to low-level attributes such as clarity and lighting. Specifically, we construct the LLSAVisionQA dataset, comprising 2,990 single images and 1,999 image pairs, each accompanied by an open-ended question about its low-level features. Through the evaluation of 15 MLLMs, we demonstrate that while some models exhibit robust low-level visual capabilities, their self-awareness remains relatively underdeveloped. Notably, for the same model, simpler questions are often answered more accurately than complex ones. However, self-awareness appears to improve when addressing more challenging questions. We hope that our benchmark will motivate further research, particularly focused on enhancing the self-awareness of MLLMs in tasks involving low-level visual perception and understanding.

Existing methods have achieved remarkable performance in image dehazing, particularly on synthetic datasets. However, they often struggle with real-world hazy images due to domain shift, limiting their practical applicability. This paper introduces HazeCLIP, a language-guided adaptation framework designed to enhance the real-world performance of pre-trained dehazing networks. Inspired by the Contrastive Language-Image Pre-training (CLIP) model’s ability to distinguish between hazy and clean images, we leverage it to evaluate dehazing results. Combined with a region-specific dehazing technique and tailored prompt sets, the CLIP model accurately identifies hazy areas, providing a high-quality, human-like prior that guides the fine-tuning process of pre-trained networks. Extensive experiments demonstrate that HazeCLIP achieves state-of-the-art performance in real-word image dehazing, evaluated through both visual quality and image quality assessment metrics. Codes are available at https://github.com/Troivyn/HazeCLIP.

Unsupervised visual representation learning has gained significant attention in the computer vision community, driven by recent advancements in contrastive learning. Most existing contrastive learning frameworks rely on instance discrimination as a pretext task, treating each instance as a distinct category. However, this often leads to intra-class collision in a large latent space, compromising the quality of learned representations. To address this issue, we propose a novel contrastive learning method that utilizes randomly generated supervision signals. Our framework incorporates two projection heads: one handles conventional classification tasks, while the other employs a random algorithm to generate fixed-length vectors representing different classes. The second head executes a supervised contrastive learning task based on these vectors, effectively clustering instances of the same class and increasing the separation between different classes. Our method, Contrastive Learning via Randomly Generated Supervision(CLRGS), significantly improves the quality of feature representations across various datasets and achieves state-of-the-art performance in contrastive learning tasks.

Although 3D generated content (3DGC) offers advantages in reducing production costs and accelerating design timelines, its quality often falls short when compared to 3D professionally generated content. Common quality issues frequently affect 3DGC, highlighting the importance of timely and effective quality assessment. Such evaluations not only ensure a higher standard of 3DGCs for end-users but also provide critical insights for advancing generative technologies. To address existing gaps in this domain, this paper introduces a novel 3DGC quality assessment dataset, 3DGCQA, built using 7 representative Text-to-3D generation methods. During the dataset’s construction, 50 fixed prompts are utilized to generate contents across all methods, resulting in the creation of 313 textured meshes that constitute the 3DGCQA dataset. The visualization intuitively reveals the presence of 6 common distortion categories in the generated 3DGCs. To further explore the quality of the 3DGCs, subjective quality assessment is conducted by evaluators, whose ratings reveal significant variation in quality across different generation methods. Additionally, several objective quality assessment algorithms are tested on the 3DGCQA dataset. The results expose limitations in the performance of existing algorithms and underscore the need for developing more specialized quality assessment methods. To provide a valuable resource for future research and development in 3D content generation and quality assessment, the dataset has been open-sourced in https://github.com/zyj-2000/3DGCQA.

With the emergence of Multimodal Large Language Models (MLLMs), hundreds of benchmarks have been developed to ensure the reliability of MLLMs in downstream tasks. However, the evaluation mechanism itself may not be reliable. For developers of MLLMs, questions remain about which benchmark to use and whether the test results meet their requirements. Therefore, we propose a critical principle of Information Density, which examines how much insight a benchmark can provide for the development of MLLMs. We characterize it from four key dimensions: (1) Fallacy, (2) Difficulty, (3) Redundancy, (4) Diversity. Through a comprehensive analysis of more than 10,000 samples, we measured the information density of 19 MLLM benchmarks. Experiments show that using the latest benchmarks in testing can provide more insight compared to previous ones, but there is still room for improvement in their information density. We hope this principle can promote the development and application of future MLLM benchmarks. Project page: https://github.com/lcysyzxdxc/bench4bench

Designing effective embodied multi-agent systems is critical for solving complex real-world tasks across domains. Due to the complexity of multi-agent embodied systems, existing methods fail to automatically generate safe and efficient training data for such systems. To this end, we propose the concept of compositional constraints for embodied multi-agent systems, addressing the challenges arising from collaboration among embodied agents. We design various interfaces tailored to different types of constraints, enabling seamless interaction with the physical world. Leveraging compositional constraints and specifically designed interfaces, we develop an automated data collection framework for embodied multi-agent systems and introduce the first benchmark for embodied multi-agent manipulation, RoboFactory. Based on RoboFactory benchmark, we adapt and evaluate the method of imitation learning and analyzed its performance in different difficulty agent tasks. Furthermore, we explore the architectures and training strategies for multi-agent imitation learning, aiming to build safe and efficient embodied multi-agent systems.

We introduce Lumina-Image 2.0, an advanced text-to-image generation framework that achieves significant progress compared to previous work, Lumina-Next. Lumina-Image 2.0 is built upon two key principles: (1) Unification - it adopts a unified architecture (Unified Next-DiT) that treats text and image tokens as a joint sequence, enabling natural cross-modal interactions and allowing seamless task expansion. Besides, since high-quality captioners can provide semantically well-aligned text-image training pairs, we introduce a unified captioning system, Unified Captioner (UniCap), specifically designed for T2I generation tasks. UniCap excels at generating comprehensive and accurate captions, accelerating convergence and enhancing prompt adherence. (2) Efficiency - to improve the efficiency of our proposed model, we develop multi-stage progressive training strategies and introduce inference acceleration techniques without compromising image quality. Extensive evaluations on academic benchmarks and public text-to-image arenas show that Lumina-Image 2.0 delivers strong performances even with only 2.6B parameters, highlighting its scalability and design efficiency. We have released our training details, code, and models at https://github.com/Alpha-VLLM/Lumina-Image-2.0.

Speech-driven methods for portraits are figuratively known as "Talkers" because of their capability to synthesize speaking mouth shapes and facial movements. Especially with the rapid development of the Text-to-Image (T2I) models, AI-Generated Talking Heads (AGTHs) have gradually become an emerging digital human media. However, challenges persist regarding the quality of these talkers and AGTHs they generate, and comprehensive studies addressing these issues remain limited. To address this gap, this paper presents the largest AGTH quality assessment dataset THQA-10K to date, which selects 12 prominent T2I models and 14 advanced talkers to generate AGTHs for 14 prompts. After excluding instances where AGTH generation is unsuccessful, the THQA-10K dataset contains 10,457 AGTHs. Then, volunteers are recruited to subjectively rate the AGTHs and give the corresponding distortion categories. In our analysis for subjective experimental results, we evaluate the performance of talkers in terms of generalizability and quality, and also expose the distortions of existing AGTHs. Finally, an objective quality assessment method based on the first frame, Y-T slice and tone-lip consistency is proposed. Experimental results show that this method can achieve state-of-the-art (SOTA) performance in AGTH quality assessment. The work is released at https://github.com/zyj-2000/Talker.

How to accurately and efficiently assess AI-generated images (AIGIs) remains a critical challenge for generative models. Given the high costs and extensive time commitments required for user studies, many researchers have turned towards employing large multi-modal models (LMMs) as AIGI evaluators, the precision and validity of which are still questionable. Furthermore, traditional benchmarks often utilize mostly natural-captured content rather than AIGIs to test the abilities of LMMs, leading to a noticeable gap for AIGIs. Therefore, we introduce **A-Bench** in this paper, a benchmark designed to diagnose *whether LMMs are masters at evaluating AIGIs*. Specifically, **A-Bench** is organized under two key principles: 1) Emphasizing both high-level semantic understanding and low-level visual quality perception to address the intricate demands of AIGIs. 2) Various generative models are utilized for AIGI creation, and various LMMs are employed for evaluation, which ensures a comprehensive validation scope. Ultimately, 2,864 AIGIs from 16 text-to-image models are sampled, each paired with question-answers annotated by human experts. We hope that **A-Bench** will significantly enhance the evaluation process and promote the generation quality for AIGIs.

In recent years, significant efforts have been dedicated to advancing 3D content generation. However, existing quality assessment research predominantly focuses on evaluating Text-to-3D Content (T23DC) while ignoring Single Image-to-3D Content (SI23DC). In this paper, we establish the first Single Image-to-3D Content Quality Assessment (SI23DCQA) database to comprehensively study the perceptual quality of SI23DCs. The database contains 1500 SI23DCs, which are generated by 5 common SI23DC algorithms from 300 images including realistic images, AI generated images, and model rendered images. Afterward, we carry out a well-designed subjective experiment to collect subjective quality ratings for SI23DCs from three perspectives including overall, color, and shape. Additionally, a benchmark experiment is conducted with the state-of-the-art no reference image quality assessment (NR-IQA), no reference video quality assessment (NR-VQA), and no reference 3D quality assessment (NR-3DQA) and the experimental results show that current quality assessment methods are limited in evaluating the perceptual loss of SI23DCs. The database is released on https://github.com/ZedFu/SI23DCQA.

Point clouds often suffer from geometric and color noise, as well as compression artifacts, during their production, storage, and transmission. Therefore, accurately and automatically evaluating the quality of point clouds is crucial for optimizing storage and compression strategies. This paper introduces the VIP-PCQA, a novel framework that combines Video, Image, and Point cloud modalities for no-reference Point Cloud Quality Assessment. The framework begins by rendering projection videos and normal images from point clouds, followed by sampling patches and computing statistical features related to color and geometry. Subsequently, a video encoder, two image encoders, and a point cloud encoder are employed to extract modality-specific features. Finally, these features are fused to regress the quality score. Experimental results on three publicly available benchmark databases demonstrate that VIP-PCQA achieves outstanding performance with excellent generalization capabilities. An ablation study further highlights the indispensable contribution of each modality to the framework’s success. The code is released on https://github.com/ZedFu/VIP-PCQA.

Current Visual Language Models (VLMs) show impressive image understanding but struggle with visual illusions, especially in real-world scenarios. Existing benchmarks focus on classical cognitive illusions, which have been learned by state-of-the-art (SOTA) VLMs, revealing issues such as hallucinations and limited perceptual abilities. To address this gap, we introduce IllusionBench, a comprehensive visual illusion dataset that encompasses not only classic cognitive illusions but also real-world scene illusions. This dataset features 1,051 images, 5,548 question-answer pairs, and 1,051 golden text descriptions that address the presence, causes, and content of the illusions. We evaluate ten SOTA VLMs on this dataset using true-or-false, multiple-choice, and open-ended tasks. In addition to real-world illusions, we design trap illusions that resemble classical patterns but differ in reality, highlighting hallucination issues in SOTA models. The top-performing model, GPT-4o, achieves 80.59% accuracy on true-or-false tasks and 76.75% on multiple-choice questions, but still lags behind human performance. In the semantic description task, GPT-4o’s hallucinations on classical illusions result in low scores for trap illusions, even falling behind some open-source models. IllusionBench is, to the best of our knowledge, the largest and most comprehensive benchmark for visual illusions in VLMs to date.

The advent of generative AI has revolutionized 3D content design, significantly enhancing modelers’ efficiency. However, the quality of generated 3D content, particularly Generated Meshes (GMs), remains a critical concern. GMs pose unique challenges for quality assessment due to their complex geometry, detailed texture mapping, and distortions that differ from traditional meshes. Existing methods fail to address these GM-specific issues. To tackle this gap, we introduce a novel no-reference quality assessment method, CAP, which integrates CT-Slice, prompt Alignment, and Projections. CAP employs a six-face projection to capture external features and a CT-like slicing approach to extract internal quality features. Additionally, it leverages Contrastive Language-Image Pre-Training (CLIP) to measure the alignment between projection embeddings and prompts as a key quality indicator. Experimental results demonstrate that CAP effectively evaluates GM quality by combining internal, external, and alignment features. The code for this work has been open-sourced in https://github.com/zyj-2000/CAP.

The advent and proliferation of large multi-modal models (LMMs) have introduced new paradigms to computer vision, transforming various tasks into a unified visual question answering framework. Video Quality Assessment (VQA), a classic field in low-level visual perception, focused initially on quantitative video quality scoring. However, driven by advances in LMMs, it is now progressing toward more holistic visual quality understanding tasks. Recent studies in the image domain have demonstrated that Visual Question Answering (VQA) can markedly enhance low-level visual quality evaluation. Nevertheless, related work has not been explored in the video domain, leaving substantial room for improvement. To address this gap, we introduce the VQA² Instruction Dataset-the first visual question answering instruction dataset that focuses on video quality assessment. This dataset consists of 3 subsets and covers various video types, containing 157,755 instruction question-answer pairs. Then, leveraging this foundation, we present the VQA² series models. The VQA² series models interleave visual and motion tokens to enhance the perception of spatial-temporal quality details in videos. We conduct extensive experiments on video quality scoring and understanding tasks, and results demonstrate that the VQA² series models achieve excellent performance in both tasks. Notably, our final model, the VQA²-Assistant, exceeds the renowned GPT-4o in visual quality understanding tasks while maintaining strong competitiveness in quality scoring tasks. Our work provides a foundation and feasible approach for integrating low-level video quality assessment and understanding with LMMs.

Ultra-low bitrate image compression is a challenging and demand- ing topic. With the development of Large Multimodal Models (LMMs), a Cross Modality Compression (CMC) paradigm of Image-Text- Image has emerged. Compared with traditional codecs, this semantic- level compression can reduce image data size to 0.1% or even lower, which has strong potential applications. However, CMC has cer- tain defects in consistency with the original image and perceptual quality. To inspire insights into such a problem, we introduce CMC- Bench, a benchmark of the cooperative performance of Image-to- Text (I2T) and Text-to-Image (T2I) models for image compression. This benchmark covers 18,000 and 40,000 images respectively to verify 6 mainstream I2T and 12 T2I models, including 160,000 sub- jective preference scores annotated by human experts. At ultra-low bitrates, it proves that the combination of some I2T and T2I models has surpassed the most advanced visual signal codecs; meanwhile, it highlights where LMMs can be further optimized toward the compression task. We encourage LMM developers to participate in this test to promote the evolution of visual signal codec protocols.

AI-based image enhancement techniques have been widely adopted in various visual applications, significantly improving the perceptual quality of user-generated content (UGC). However, the lack of specialized quality assessment models has become a significant limiting factor in this field, limiting user experience and hindering the advancement of enhancement methods. While perceptual quality assessment methods have shown strong performance on UGC and AIGC individually, their effectiveness on AI-enhanced UGC (AI-UGC) which blends features from both-remains largely unexplored. To address this gap, we construct AU-IQA, a benchmark dataset comprising 4,800 AI-UGC images produced by three representative enhancement types which include super-resolution, low-light enhancement, and denoising. On this dataset, we further evaluate a range of existing quality assessment models, including traditional IQA methods and large multimodal models. Finally, we provide a comprehensive analysis of how well current approaches perform in assessing the perceptual quality of AI-UGC. The access link to the AU-IQA is https://github.com/WNNGGU/AU-IQA-Dataset.

International audience

360° omnidirectional images (ODIs) have gained considerable attention recently, and are widely used in various virtual reality (VR) and augmented reality (AR) applications. However, capturing such images is expensive and requires specialized equipment, making ODI synthesis increasingly important. While common 2D image generation and editing methods are rapidly advancing, these models struggle to deliver satisfactory results when generating or editing ODIs due to the unique format and broad 360° Field-of-View (FoV) of ODIs. To bridge this gap, we construct Any2Omni , the first comprehensive ODI generation-editing dataset comprises 60,000+ training data covering diverse input conditions and up to 9 ODI generation and editing tasks. Built upon Any2Omni, we propose an Omni model for Omni-directional image generation and editing ( Omni 2), with the capability of handling various ODI generation and editing tasks under diverse input conditions using one model. Extensive experiments demonstrate the superiority and effectiveness of the proposed Omni2 model for both the ODI generation and editing tasks. Both the Any2Omni dataset and the Omni2 model are publicly available at: https://github.com/IntMeGroup/Omni2.

Coordinating multiple embodied agents in dynamic environments remains a core challenge in artificial intelligence, requiring both perception-driven reasoning and scalable cooperation strategies. While recent works have leveraged large language models (LLMs) for multi-agent planning, a few have begun to explore vision-language models (VLMs) for visual reasoning. However, these VLM-based approaches remain limited in their support for diverse embodiment types. In this work, we introduce VIKI-Bench, the first hierarchical benchmark tailored for embodied multi-agent cooperation, featuring three structured levels: agent activation, task planning, and trajectory perception. VIKI-Bench includes diverse robot embodiments, multi-view visual observations, and structured supervision signals to evaluate reasoning grounded in visual inputs. To demonstrate the utility of VIKI-Bench, we propose VIKI-R, a two-stage framework that fine-tunes a pretrained vision-language model (VLM) using Chain-of-Thought annotated demonstrations, followed by reinforcement learning under multi-level reward signals. Our extensive experiments show that VIKI-R significantly outperforms baselines method across all task levels. Furthermore, we show that reinforcement learning enables the emergence of compositional cooperation patterns among heterogeneous agents. Together, VIKI-Bench and VIKI-R offer a unified testbed and method for advancing multi-agent, visual-driven cooperation in embodied AI systems.

Video generation has achieved significant advances through rectified flow techniques, but issues like unsmooth motion and misalignment between videos and prompts persist. In this work, we develop a systematic pipeline that harnesses human feedback to mitigate these problems and refine the video generation model. Specifically, we begin by constructing a large-scale human preference dataset focused on modern video generation models, incorporating pairwise annotations across multi-dimensions. We then introduce VideoReward, a multi-dimensional video reward model, and examine how annotations and various design choices impact its rewarding efficacy. From a unified reinforcement learning perspective aimed at maximizing reward with KL regularization, we introduce three alignment algorithms for flow-based models. These include two training-time strategies: direct preference optimization for flow (Flow-DPO) and reward weighted regression for flow (Flow-RWR), and an inference-time technique, Flow-NRG, which applies reward guidance directly to noisy videos. Experimental results indicate that VideoReward significantly outperforms existing reward models, and Flow-DPO demonstrates superior performance compared to both Flow-RWR and supervised fine-tuning methods. Additionally, Flow-NRG lets users assign custom weights to multiple objectives during inference, meeting personalized video quality needs.

Animated QR codes present an exciting frontier for dynamic content delivery and digital interaction. However, despite their potential, there has been no prior work focusing on the generation of animated QR codes that are both visually appealing and universally scannable. In this paper, we introduce AnimateQR, **the first generative framework** for creating **animated QR codes** that balance aesthetic flexibility with scannability. Unlike previous methods that focus on static QR codes, AnimateQR leverages **hierarchical luminance guidance** and **progressive spatiotemporal control** to produce high-quality dynamic QR codes. Our first innovation is a multi-scale hierarchical control signal that adjusts luminance across different spatial scales, ensuring that the QR code remains decodable while allowing for artistic expression. The second innovation is a progressive control mechanism that dynamically adjusts spatiotemporal guidance throughout the diffusion denoising steps, enabling fine-grained balance between visual quality and scannability. Extensive experimental results demonstrate that AnimateQR achieves state-of-the-art performance in both decoding success rates (96\% vs. 56\% baseline) and visual quality (user preference: 7.2 vs. 2.3 on a 10-point scale). Codes are availble at https://github.com/mulns/AnimateQR.

Recent image quality assessment (IQA) methods typically focus on predicting the mean opinion score (MOS) of image quality, ignoring the image quality score distribution. This distribution provides valuable information beyond the MOS, including the standard deviation of opinion scores (SOS) and opinion scores at different quality levels. This paper introduces a novel no-reference IQA method that predicts the image quality score distribution to estimate the MOS. The proposed method consists of three modules: a visual feature extraction module, a graph convolutional module, and a MOS prediction module. In the visual feature extraction module, a convolutional neural network is designed to extract both first- and second-order visual features of images. The graph convolutional module employs a graph convolutional network (GCN)-based mapper to map these visual features to the image quality score distribution by exploring correlations between quality labels. The MOS is then derived from the predicted image quality score distribution in the MOS prediction module. We are the first to jointly train the method using both the MOS and the image quality score distribution, enabling it to learn richer subjective information and improve prediction performance. To address the lack of the ground-truth image quality score distribution in some IQA databases, we propose to use a SOS assumption to generate a Gaussian-based image quality score distribution that better reflects subjective perception. Additionally, we design appropriate loss functions for training. Experimental results demonstrate that our method effectively predicts both the image quality score distribution and the MOS, outperforming most state-of-the-art IQA methods.

Video frame interpolation (VFI) synthesizes new frames from original video frames to produce high frame-rate videos and enhance their visual appeal. The quality of these interpolated frames significantly affects the perceptual experience of the synthesized video. Recent research in VFI has increasingly focused on perceptual quality of the interpolated frames and the overall video. However, most existing quality metrics do not align well with human perceptual experiences and often suffer from unnatural artifacts in the interpolated frames. Consequently, there is an urgent need for VFI video quality assessment (VFIVQA) methods to assess the quality of the synthesized videos. In this paper, we propose both a full-reference (FR) method and a no-reference (NR) method for VFIVQA. The FR method employs two feature extraction blocks to measure continuous frame changes, extracting flow features with short temporal spans and motion features with long temporal spans. By calculating multilevel similarities in the temporal dimension of 3D convolutional neural networks and fusing these similarity features, the quality score of the VFI video is obtained from the quality regression network. Since the flow feature extraction block does not utilize the reference VFI video, the proposed NR method consists solely of this feature block. Extensive validation on several VFIVQA datasets demonstrates that the proposed methods outperform state-of-the-art FR and NR methods.

Animated human (AH) have gained popularity due to their vivid appearance and smooth, natural movements. Various animation methods based on artificial intelligence (AI) have been introduced, which are viewed as “Imitators,” offering new solutions for designing AHs. However, the effectiveness of these AI-generated AHs varies significantly across different categories and within the same category, leading to visual distortions that adversely affect the viewer’s experience. Consequently, it is essential to evaluate the quality of AHs to provide reliable and objective indicators for their further development and to ensure the delivery of higher-quality AH videos to users. In this paper, the first Animated Human Quality Assessment (AHQA) dataset is constructed by selecting 6 advanced and popular imitators and 10 common actions to animate 20 AI-generated characters. The constructed dataset integrates different genders and age groups of character images, and two types of poses, standing and sitting, are selected, highlighting the comprehensiveness and diversity of the AHQA dataset. Subjective experiments reveal significant differences in the quality of AHs produced by different imitators. Finally, we propose a quality assessment method, VIP-QA, incorporating Video quality, Identity consistency, and Posture similarity for the AHQA dataset. Experimental results show that VIP-QA significantly outperforms existing assessment methods on multiple datasets by about 5%, more closely approximates human visual perception, and provides a valid objective metric for assessing imitators. All the work in this paper has been released at https://github.com/zyj-2000/Imitator.

With the evolution of storage and communication protocols, ultra-low bitrate image compression has become a highly demanding topic. However, all existing compression algorithms must sacrifice either consistency with the ground truth or perceptual quality at ultra-low bitrate. During recent years, the rapid development of the Large Multimodal Model (LMM) has made it possible to balance these two goals. To solve this problem, this paper proposes a method called Multimodal Image Semantic Compression (MISC), which consists of an LMM encoder for extracting the semantic information of the image, a map encoder to locate the region corresponding to the semantic, an image encoder generates an extremely compressed bitstream, and a decoder reconstructs the image based on the above information. Experimental results show that our proposed MISC is suitable for compressing both traditional Natural Sense Images (NSIs) and emerging AI-Generated Images (AIGIs) content. It can achieve optimal consistency and perception results while saving 50% bitrate, which has strong potential applications in the next generation of storage and communication. The code will be released on https://github.com/lcysyzxdxc/MISC.

Digital humans have witnessed extensive applications in various domains, necessitating related quality assessment studies. However, there is a lack of comprehensive digital human quality assessment (DHQA) databases. To address this gap, we propose SJTU-H3D, a subjective quality assessment database specifically designed for full-body digital humans. It comprises 40 high-quality reference digital humans and 1,120 labeled distorted counterparts generated with seven types of distortions. The SJTU-H3D database can serve as a benchmark for DHQA research, allowing evaluation and refinement of processing algorithms. Further, we propose a zero-shot DHQA approach that focuses on no-reference (NR) scenarios to ensure generalization capabilities while mitigating database bias. Our method leverages semantic and distortion features extracted from projections, as well as geometry features derived from the mesh structure of digital humans. Specifically, we employ the Contrastive Language-Image Pre-training (CLIP) model to measure semantic affinity and incorporate the Naturalness Image Quality Evaluator (NIQE) model to capture low-level distortion information. Additionally, we utilize dihedral angles as geometry descriptors to extract mesh features. By aggregating these measures, we introduce the Digital Human Quality Index (DHQI), which demonstrates significant improvements in zero-shot performance. The DHQI can also serve as a robust baseline for DHQA tasks, facilitating advancements in the field. The database and the code are available at https://github.com/zzc-1998/SJTU-H3D.

Recent advancements in text-to-image (T2I) generation have spurred the development of text-to-3D asset (T23DA) generation, leveraging pretrained 2D text-to-image diffusion models for text-to-3D asset synthesis. Despite the growing popularity of text-to-3D asset generation, its evaluation has not been well considered and studied. However, given the significant quality discrepancies among various text-to-3D assets, there is a pressing need for quality assessment models aligned with human subjective judgments. To tackle this challenge, we conduct a comprehensive study to explore the T23DA quality assessment (T23DAQA) problem in this work from both subjective and objective perspectives. Given the absence of corresponding databases, we first establish the largest text-to-3D asset quality assessment database to date, termed the AIGC-T23DAQA database. This database encompasses 969 validated 3D assets generated from 170 prompts via 6 popular text-to-3D asset generation models, and corresponding subjective quality ratings for these assets from the perspectives of quality, authenticity, and text-asset correspondence, respectively. Subsequently, we establish a comprehensive benchmark based on the AIGC-T23DAQA database, and devise an effective T23DAQA model to evaluate the generated 3D assets from the aforementioned three perspectives, respectively. Specifically, the proposed method utilizes the projection videos of text-to-3D assets to extract 3D shape, texture and text-asset correspondence features, then fuses them to calculate the final three preference scores respectively. Extensive experimental results demonstrate the effectiveness of the proposed T23DAQA method in evaluating the quality of AI generated 3D asset, which is more consistent with human perception. To the best of our knowledge, this is the first work that studies the problem of text-guided 3D generation quality assessment, and our database and codes will be released to facilitate future research.

The importance of visual quality in point clouds has been significantly underlined due to the rapid rise in 3D vision applications which aim to deliver affordable and superior user experiences. Reviewing the evolution of point cloud quality assessment (PCQA), it’s observed that visual quality evaluation typically employs single-modal data, either sourced from 2D projections or the 3D point clouds. The 2D projections possess abundant texture and semantic information while they are heavily reliant on viewpoints. In contrast, 3D point clouds are more reactive to geometric distortions and viewpoint-invariant. Consequently, to maximize the benefits of both point cloud and image modalities, we present an advanced no-reference Multi-Modal Point Cloud Quality Assessment (MM-PCQA+) metric. Specifically, we divide the point clouds into sub-models to reflect local geometric distortions such as point shifting and down-sampling. Afterwards, we render the point clouds using a cube-like projection setup and sample the projections of interest using a point-visible-ratio for image feature extraction. In order to fulfill these objectives, the sub-models and projected images are encoded using point-based and image-based neural networks. Lastly, we implement symmetric cross-modal attention to amalgamate multi-modal quality-aware features. Experimental results demonstrate that our metric surpasses all state-of-the-art methods and significantly advances beyond previous no-reference PCQA methods.

Quality assessment, which evaluates the visual quality level of multimedia experiences, has garnered significant attention from researchers and has evolved substantially through dedicated efforts. Before the advent of large models, quality assessment typically relied on small expert models tailored for specific tasks. While these smaller models are effective at handling their designated tasks and predicting quality levels, they often lack explainability and robustness. With the advancement of large models, which align more closely with human cognitive and perceptual processes, many researchers are now leveraging the prior knowledge embedded in these large models for quality assessment tasks. This emergence of quality assessment within the context of large models motivates us to provide a comprehensive review focusing on two key aspects: (1) the assessment of large models and (2) the role of large models in assessment tasks. We begin by reflecting on the historical development of quality assessment. Subsequently, we move to detailed discussions of related works concerning quality assessment in the era of large models. Finally, we offer insights into the future progression and potential pathways for quality assessment in this new era. We hope that this survey will enable a rapid understanding of the development of quality assessment in the era of large models and inspire further advancements in the field.

The Segment Anything Model (SAM) has recently gained popularity in the filed of 3D medical segmentation for its remarkable generalization capability. However, most SAM-based 2D to 3D adaptation methods requires prompting each slice in a medical volume, which is label-consuming. How to effectively propagate the prompts on the single slice to all slices while ensuring the continuity of the entire volume boundaries is still an unresolved challenge. To address this, we propose Adjacent-Slice Correspondence-guided 3D SAM(ASC-3DSAM), a novel approach for 3D medical segmentation, which matches key points or features between adjacent slices to establish correspondences for neighboring slices, thereby propagating prompt information to the entire volume using just a single prompted slice. Specifically, we present a Slice-Correlation Module (SCM), the adjacent slice as the query is segmented using the mask information from the central slice, thereby propagating prompt information to the entire volume. In addition, a hybrid continuity loss called Continuous Boundary-aware Loss(CBL) is proposed, which takes the form of a distance metric on the space of contours and integrates gradient information of the volume boundaries to supervise the continuity of the entire volume. Our approach is evaluated on two public datasets, including AMOS and BTCV. Experimental results demonstrate that ASC-3DSAM achieves state-of-the-art (SOTA) performance, with an improvement of approximately 1.86% of Dice and 1.92(mm) of HD95 compared to the next-best method. Furthermore, our method can produce smooth closed boundaries of the entire volumes outperforming the SOTA models.

Previous methods for Video Frame Interpolation (VFI) have encountered challenges, notably the manifestation of blur and ghosting effects. These issues can be traced back to two pivotal factors: unavoidable motion errors and misalignment in supervision. In practice, motion estimates often prove to be error-prone, resulting in misaligned features. Furthermore, the reconstruction loss tends to bring blurry results, particularly in misaligned regions. To mitigate these challenges, we propose a new paradigm called PerVFI (Perception-oriented Video Frame Interpolation). Our approach incorporates an Asymmetric Synergistic Blending module (ASB) that utilizes features from both sides to synergistically blend intermediate features. One reference frame emphasizes primary content, while the other contributes complementary information. To impose a stringent constraint on the blending process, we introduce a self-learned sparse quasi-binary mask which effectively mitigates ghosting and blur artifacts in the output. Additionally, we employ a normalizing flow-based generator and utilize the negative log-likelihood loss to learn the conditional distribution of the output, which further facilitates the generation of clear and fine details. Experimental results validate the superiority of PerVFI, demonstrating significant improvements in perceptual quality compared to existing methods. Codes are available at https://github.com/mulns/PerVFI

In the digital era, QR codes serve as a linchpin connecting virtual and physical realms. Their pervasive integration across various applications highlights the demand for aesthetically pleasing codes without compromised scannability. However, prevailing methods grapple with the intrinsic challenge of balancing customization and scannability. Notably, stable-diffusion models have ushered in an epoch of high-quality, customizable content generation. This paper introduces Text2QR, a pioneering approach leveraging these advancements to address a fundamental challenge: concurrently achieving user-defined aesthetics and scanning robustness. To ensure stable generation of aesthetic QR codes, we introduce the QR Aesthetic Blueprint (QAB) module, generating a blueprint image exerting control over the entire generation process. Subsequently, the Scannability Enhancing Latent Refinement (SELR) process refines the output iteratively in the latent space, enhancing scanning robustness. This approach harnesses the potent generation capabilities of stable-diffusion models, navigating the trade-off between image aesthetics and QR code scannability. Our experiments demonstrate the seamless fusion of visual appeal with the practical utility of aesthetic QR codes, markedly outperforming prior methods. Codes are available at https://github.com/mulns/Text2QR

Recently, many algorithms have employed image-adaptive lookup tables (LUTs) to achieve real-time image enhancement. Nonetheless, a prevailing trend among existing methods has been the employment of linear combinations of basic LUTs to formulate image-adaptive LUTs, which limits the generalization ability of these methods. To address this limitation, we propose a novel framework named AttentionLut for real-time image enhancement, which utilizes the attention mechanism to generate image-adaptive LUTs. Our proposed framework consists of three lightweight modules. We begin by employing the global image context feature module to extract image-adaptive features. Subsequently, the attention fusion module integrates the image feature with the priori attention feature obtained during training to generate image-adaptive canonical polyadic tensors. Finally, the canonical polyadic reconstruction module is deployed to reconstruct image-adaptive residual 3DLUT, which is subsequently utilized for enhancing input images. Experiments on the benchmark MIT-Adobe FiveK dataset demonstrate that the proposed method achieves better enhancement performance quantitatively and qualitatively than the state-of-the-art methods.

Although current data augmentation methods are successful to alleviate the data insufficiency, conventional augmentation are primarily intra-domain while advanced generative adversarial networks (GANs) generate images remaining uncertain, particularly in small-scale datasets. In this paper, we propose a parameterized GAN (ParaGAN) that effectively controls the changes of synthetic samples among domains and highlights the attention regions for downstream classification. Specifically, ParaGAN incorporates projection distance parameters in cyclic projection and projects the source images to the decision boundary to obtain the class-difference maps. Our experiments show that ParaGAN can consistently outperform the existing augmentation methods with explainable classification on two small-scale medical datasets.

In an era where 3D Digital Humans (DHs) are becoming increasingly prevalent in fields like gaming, automotive, and the metaverse, the demand for high DH visual quality is rising. This paper presents the first-ever reduced-reference (RR) quality assessment metric tailored specifically for textured mesh DHs, aiming to optimize transmission systems and improve Quality of Experience (QoE) for viewers in resource-constrained environments. Four critical geometric curvature-related attributes and two texture-related indicators are computed, which are then statistically analyzed and utilized in a Support Vector Regression (SVR) model for robust and efficient quality prediction. Experimental results confirm that our method outperforms existing full-reference (FR) metrics, making it an invaluable tool for the future of 3D DHs in various applications. The code is available at https://github.com/zzc-1998/RR-DHQA.

[?]In recent years, the rapid advancement of Artificial Intelligence Generated Content (AIGC) has attracted widespread attention. Among the AIGC, AI generated omnidirectional images hold significant potential for Virtual Reality (VR) and Augmented Reality (AR) applications, hence omnidirectional AIGC techniques have also been widely studied. AI-generated omnidirectional images exhibit unique distortions compared to natural omnidirectional images, however, there is no dedicated Image Quality Assessment (IQA) criteria for assessing them. This study addresses this gap by establishing a large-scale AI generated omnidirectional image IQA database named AIGCOIQA2024 and constructing a comprehensive benchmark. We first generate 300 omnidirectional images based on 5 AIGC models utilizing 25 text prompts. A subjective IQA experiment is conducted subsequently to assess human visual preferences from three perspectives including quality, comfortability, and correspondence. Finally, we conduct a benchmark experiment to evaluate the performance of state-of-the-art IQA models on our database. The AIGCOIQA2024 database is released to facilitate future research on https://github.com/IntMeGroup/AIGCOIQA.

In the realm of media technology, digital humans have gained prominence due to rapid advancements in computer technology. However, the manual modeling and control required for the majority of digital humans pose significant obstacles to efficient development. The speech-driven methods offer a novel avenue for manipulating the mouth shape and expressions of digital humans. Despite the proliferation of driving methods, the quality of many generated talking head (TH) videos remains a concern, impacting user visual experiences. To tackle this issue, this paper introduces the Talking Head Quality Assessment (THQA) database, featuring 800 TH videos generated through 8 diverse speechdriven methods. Extensive experiments affirm the THQA database’s richness in character and speech features. Subsequent subjective quality assessment experiments analyze correlations between scoring results and speech-driven methods, ages, and genders. In addition, experimental results show that mainstream image and video quality assessment methods have limitations for the THQA database, underscoring the imperative for further research to enhance TH video quality assessment. The THQA database is publicly accessible at https://github.com/zyj-2000/THQA.

With the rapid evolution of the Text-to-Image (T2I) model in recent years, their unsatisfactory generation result has become a challenge. However, uniformly refining AI-Generated Images (AIGIs) of different qualities not only limited optimization capabilities for low-quality AIGIs but also brought negative optimization to high-quality AIGIs. To address this issue, a quality-award refiner named Q-Refine is proposed. Based on the preference of the Human Visual System (HVS), Q-Refine uses the Image Quality Assessment (IQA) metric to guide the refining process for the first time, and modify images of different qualities through three adaptive pipelines. Experimental data shows that for mainstream T2I models, Q-Refine can perform effective optimization to AIGIs of different qualities. It can be a general refiner to optimize AIGIs from both fidelity and aesthetic quality levels, thus expanding the application of the T2I generation models. The code is released on https://github.com/Q-Future/Q-Refine.

Viewpoint selection plays a pivotal role in projection-based point cloud quality assessment (PCQA). Generally speaking, sole reliance on a single projection fails to capture adequate quality information, leading to the prevalent use of multi-projection approaches. It is important to recognize that viewpoint selection is significantly influenced by human preferences and viewpoints that align with human predilections exert a greater impact on PCQA. Therefore, we introduce the first viewpoint selection database for PCQA, which comprises 405 distorted point clouds, accompanied by preferred viewpoints collected from humans. Then we propose a novel human preference index, devised from the Visible-Points Ratio and Visible-Color-Entropy Ratio, to guide the selection of viewpoints. Our experimental findings confirm that this human preference index correlates more closely with human preferences than traditional viewpoint selection settings. Moreover, the proposed PCQA method optimized with the human preference index demonstrates competitive performance as well.

Recently, the color difference (CD) of standard dynamic range (SDR) images has attracted the attention of researchers. It is worth noting that due to the development of high dynamic range (HDR) image generation technology, the CD of the SDR and HDR images is also worth in-depth research. This is because the HDR image generated from an original SDR image may have changes in color. Some color changes can give people a comfortable impression, but this may also change the information originally expressed in the SDR image. Therefore, this paper researches the CD of the original SDR image and the generated HDR image, and proposes a network to predict the CDs of SDR and HDR image pairs. Specifically, we first build a SDR-HDR image CD dataset. The dataset contains 504 SDR and HDR image pairs, where HDR images are generated from the SDR images using five HDR image generation methods. Second, we propose a siamese neural network to predict the CDs of SDR and HDR image pairs, which consists of three parts: space conversion, feature extraction, and CD calculation. Finally, experiments prove that the proposed network has a superior ability to predict the CDs of SDR and HDR image pairs.

Despite recent advancements in generative models, the variation in image quality remains a significant concern. To tackle this issue, we propose PrefIQA, an effective human preference learning metric, which can better evaluate the quality of AI-generated images. PrefIQA consists of two units, namely Feature Extraction Unit and Feature Fusion Unit. In Feature Extraction Unit, we introduce a prompt-segmentation module to divide prompts into multiple phrases, enabling a more detailed evaluation of the alignment between images and texts. In Feature Fusion Unit, we introduce a modality-fusion module, which effectively mixes text features and image features to improve the overall performance. In the experiment part, extensive experiments are conducted, demonstrating that PrefIQA surpasses existing text-to-image alignment metrics. We believe that PrefIQA’s proposal would facilitate researches on AI-generated image quality assessment, and make a valuable contribution to the field of text-to-image generation.

In immersive multimedia systems, the perceptual quality model of omnidirectional video is indispensable. However, to cope with its resolution that is several times higher than ordinary video, the existing omnidirectional video quality assessment (OVQA) models require extremely high computational complexity and usually need to transcode the projection into a certain format. Therefore, to assess the perceptual quality of omnidirectional video effectively, we propose Projection-Aware Patch Sampling (PAPS)-OVQA to process its three common projection formats simultaneously while resizing high-resolution video into patches sampled from uniform grids and finally apply Fragment Attention Network (FANet) to perform quality regression. As a result, we avoid the overhead computational cost of projection transcoding and reduce the complexity of the quality model greatly. Experimental data show that PAPS-OVQA guarantees good performance while retaining high efficiency under different projection formats.

Text-to-image (T2I) generation is a pivotal and core interest within the realm of AI content generation. Amid the swift advancements of both open-source (such as Stable Diffusion) and proprietary (for example, DALLE, MidJourney) T2I models, there is a notable absence of a comprehensive and robust quantitative framework for evaluating their output quality. Traditional methods of quality assessment overlook the textual prompts when judging images; meanwhile, the advent of large multi-modal models (LMMs) introduces the capability to incorporate text prompts in evaluations, yet the challenge of fine-tuning these models for precise T2I quality assessment remains unresolved. In our study, we introduce the T2I-Scorer, a novel two-stage training methodology aimed at fine-tuning LMMs for T2I evaluation. For the first stage, we collect 397K GPT-4V-labeled question-answer pairs related to T2I evaluation. Termed as T2I-ITD, the pseudo-labeled dataset is analyzed and examined by human, and used for instruction tuning to improve the LMM's low-level quality perception. The first stage model, T2I-Scorer-IT, has reached superior accuracy on T2I evaluation than all kinds of existing T2I metrics under zero-shot settings. For the second stage, we define an explicit multi-task training scheme to further align the LMM with human opinion scores, and the fine-tuned T2I-Scorer can reach state-of-the-art accuracy on both image quality and image-text alignment perspectives with significant improvements. We anticipate the proposed metrics can serve as a reliable metric to gauge the ability of T2I generation models in the future. We will make code, data, and weights publicly available.

With the rapid development of generative models, AI-Generated Content (AIGC) has exponentially increased in daily lives. Among them, Text-to-Video (T2V) generation has received widespread attention. Though many T2V models have been released for generating high perceptual quality videos, there is still lack of a method to evaluate the quality of these videos quantitatively. To solve this issue, we establish the largest-scale Text-to-Video Quality Assessment DataBase (T2VQA-DB) to date. The dataset is composed of 10,000 videos generated by 9 different T2V models, along with each video's corresponding mean opinion score. Based on T2VQA-DB, we propose a novel transformer-based model for subjective-aligned Text-to-Video Quality Assessment (T2VQA). The model extracts features from text-video alignment and video fidelity perspectives, then it leverages the ability of a large language model to give the prediction score. Experimental results show that T2VQA outperforms existing T2V metrics and SOTA video quality assessment models. Quantitative analysis indicates that T2VQA is capable of giving subjective-align predictions, validating its effectiveness. The dataset and code are available at https://github.com/QMME/T2VQA.

Although large multi-modality models (LMMs) have seen extensive exploration and application in various quality assessment studies, their integration into Point Cloud Quality Assessment (PCQA) remains unexplored. Given LMMs' exceptional performance and robustness in low-level vision and quality assessment tasks, this study aims to investigate the feasibility of imparting PCQA knowledge to LMMs through text supervision. To achieve this, we transform quality labels into textual descriptions during the fine-tuning phase, enabling LMMs to derive quality rating logits from 2D projections of point clouds. To compensate for the loss of perception in the 3D domain, structural features are extracted as well. These quality logits and structural features are then combined and regressed into quality scores. Our experimental results affirm the effectiveness of our approach, showcasing a novel integration of LMMs into PCQA that enhances model understanding and assessment accuracy. We hope our contributions can inspire subsequent investigations into the fusion of LMMs with PCQA, fostering advancements in 3D visual quality analysis and beyond. The code is available at https://github.com/zzc-1998/LMM-PCQA.

In recent years, immersive communication has emerged as a compelling alternative to traditional video communication methods. One prospective avenue for immersive communication involves augmenting the user's immersive experience through the transmission of three-dimensional (3D) talking heads (THs). However, transmitting 3D THs poses significant challenges due to its complex and voluminous nature, often leading to pronounced distortion and a compromised user experience. Addressing this challenge, we introduce the 3D Talking Heads Quality Assessment (THQA-3D) dataset, comprising 1,000 sets of distorted and 50 original TH mesh sequences (MSs), to facilitate quality assessment in 3D TH transmission. A subjective experiment, characterized by a novel interactive approach, is conducted with recruited participants to assess the quality of MSs in THQA-3D dataset. Leveraging this dataset, we also propose a multimodal Quality-of-Experience (QoE) method incorporating a Large Quality Model (LQM). This method involves frontal projection of MSs and subsequent rendering into videos, with quality assessment facilitated by the LQM and a variable-length video memory filter (VVMF). Additionally, tone-lip coherence and silence detection techniques are employed to characterize audio-visual coherence in 3D MS streams. Experimental evaluation demonstrates the proposed method's superiority, achieving state-of-the-art performance on the THQA-3D dataset and competitiveness on other QoE datasets. Both the THQA-3D dataset and the QoE model have been publicly released at https://github.com/zyj-2000/THQA-3D

Existing methods to generate aesthetic QR codes, such as image and style transfer techniques, tend to compromise either the visual appeal or the scannability of QR codes when they incorporate human face identity. Addressing these imperfections, we present Face2QR—a novel pipeline specifically designed for generating personalized QR codes that harmoniously blend aesthetics, face identity, and scannability. Our pipeline introduces three innovative components. First, the ID-refined QR integration (IDQR) seamlessly intertwines the background styling with face ID, utilizing a unified SD-based framework with control networks. Second, the ID-aware QR ReShuffle (IDRS) effectively rectifies the conflicts between face IDs and QR patterns, rearranging QR modules to maintain the integrity of facial features without compromising scannability. Lastly, the ID-preserved Scannability Enhancement (IDSE) markedly boosts scanning robustness through latent code optimization, striking a delicate balance between face ID, aesthetic quality and QR functionality. In comprehensive experiments, Face2QR demonstrates remarkable performance, outperforming existing approaches, particularly in preserving facial recognition features within custom QR code designs.

Exposure Correction (EC) aims to recover proper exposure conditions for images captured under over-exposure or under-exposure scenarios. While existing deep learning models have shown promising results, few have fully embedded Retinex theory into their architecture, highlighting a gap in current methodologies. Additionally, the balance between high performance and efficiency remains an under-explored problem for exposure correction task. Inspired by Mamba which demonstrates powerful and highly efficient sequence modeling, we introduce a novel framework based on \textbf{Mamba} for \textbf{E}xposure \textbf{C}orrection (\textbf{ECMamba}) with dual pathways, each dedicated to the restoration of reflectance and illumination map, respectively. Specifically, we firstly derive the Retinex theory and we train a Retinex estimator capable of mapping inputs into two intermediary spaces, each approximating the target reflectance and illumination map, respectively. This setup facilitates the refined restoration process of the subsequent \textbf{E}xposure \textbf{C}orrection \textbf{M}amba \textbf{M}odule (\textbf{ECMM}). Moreover, we develop a novel \textbf{2D S}elective \textbf{S}tate-space layer guided by \textbf{Retinex} information (\textbf{Retinex-SS2D}) as the core operator of \textbf{ECMM}. This architecture incorporates an innovative 2D scanning strategy based on deformable feature aggregation, thereby enhancing both efficiency and effectiveness. Extensive experiment results and comprehensive ablation studies demonstrate the outstanding performance and the importance of each component of our proposed ECMamba. Code is available at \url{https://github.com/LowlevelAI/ECMamba}.

Large numbers of synthesized videos from diffusion models pose threats to information security and authenticity, leading to an increasing demand for generated content detection. However, existing video-level detection algorithms primarily focus on detecting facial forgeries and often fail to identify diffusion-generated content with a diverse range of semantics. To advance the field of video forensics, we propose an innovative algorithm named Multi-Modal Detection(MM-Det) for detecting diffusion-generated videos. MM-Det utilizes the profound perceptual and comprehensive abilities of Large Multi-modal Models (LMMs) by generating a Multi-Modal Forgery Representation (MMFR) from LMM's multi-modal space, enhancing its ability to detect unseen forgery content. Besides, MM-Det leverages an In-and-Across Frame Attention (IAFA) mechanism for feature augmentation in the spatio-temporal domain. A dynamic fusion strategy helps refine forgery representations for the fusion. Moreover, we construct a comprehensive diffusion video dataset, called Diffusion Video Forensics (DVF), across a wide range of forgery videos. MM-Det achieves state-of-the-art performance in DVF, demonstrating the effectiveness of our algorithm. Both source code and DVF are available at https://github.com/SparkleXFantasy/MM-Det.

Parameter-efficient transfer learning (PETL) methods show promise in adapting a pre-trained model to various downstream tasks while training only a few parameters. In the computer vision (CV) domain, numerous PETL algorithms have been proposed, but their direct employment or comparison remains inconvenient. To address this challenge, we construct a Unified Visual PETL Benchmark (V-PETL Bench) for the CV domain by selecting 30 diverse, challenging, and comprehensive datasets from image recognition, video action recognition, and dense prediction tasks. On these datasets, we systematically evaluate 25 dominant PETL algorithms and open-source a modular and extensible codebase for fair evaluation of these algorithms. V-PETL Bench runs on NVIDIA A800 GPUs and requires approximately 310 GPU days. We release all the benchmark, making it more efficient and friendly to researchers. Additionally, V-PETL Bench will be continuously updated for new PETL algorithms and CV tasks.

With the rapid advancements of the text-to-image generative model, AI-generated images (AGIs) have been widely applied to entertainment, education, social media, etc. However, considering the large quality variance among different AGIs, there is an urgent need for quality models that are consistent with human subjective ratings. To address this issue, we extensively consider various popular AGI models, generated AGI through different prompts and model parameters, and collected subjective scores at the perceptual quality and text-to-image alignment, thus building the most comprehensive AGI subjective quality database AGIQA-3K so far. Furthermore, we conduct a benchmark experiment on this database to evaluate the consistency between the current Image Quality Assessment (IQA) model and human perception, while proposing StairReward that significantly improves the assessment performance of subjective text-to-image alignment. We believe that the fine-grained subjective scores in AGIQA-3K will inspire subsequent AGI quality models to fit human subjective perception mechanisms at both perception and alignment levels and to optimize the generation result of future AGI models. The database is released on https://github.com/lcysyzxdxc/AGIQA-3k-Database.

Online video streaming has fundamental limitations on the transmission bandwidth and computational capacity and super-resolution is a promising potential solution. However, applying existing video super-resolution methods to online streaming is non-trivial. Existing video codecs and streaming protocols (e.g., WebRTC) dynamically change the video quality both spatially and temporally, which leads to diverse and dynamic degradations. Furthermore, online streaming has a strict requirement for latency that most existing methods are less applicable. As a result, this paper focuses on the rarely exploited problem setting of online streaming video super resolution. To facilitate the research on this problem, a new benchmark dataset named LDV-WebRTC is constructed based on a real-world online streaming system. Leveraging the new benchmark dataset, we propose a novel method specifically for online video streaming, which contains a convolution and Look-Up Table (LUT) hybrid model to achieve better performance-latency trade-off. To tackle the changing degradations, we propose a mixture-of-expert-LUT module, where a set of LUT specialized in different degradations are built and adaptively combined to handle different degradations. Experiments show our method achieves 720P video SR around 100 FPS, while significantly outperforms existing LUT-based methods and offers competitive performance compared to efficient CNN-based methods. Code is available at https://github.com/quzefan/ConvLUT.

The prevalence of the Internet of Things (IoT) has led to the widespread adoption of 2D barcodes as a means of offline-to-online communication. Whereas, 2D barcodes are not ideal for publicity materials, due to their space-consuming nature. Recent works have proposed 2D image barcodes that contain invisible codes or hyperlinks to transmit hidden information from offline to online. However, these methods undermine the purpose of the codes being invisible, due to the the requirement of markers to locate them. The conference version of this work has presents a novel imperceptible information embedding framework for display or print-camera scenarios, which includes not only hiding and recvoery but also locating and correcting. With the assistance of learned invisible markers, hidden codes can be rendered truly imperceptible. A highly effective multi-stage training scheme is proposed to achieve high visual fidelity and retrieval resiliency, wherein information is concealed in a sub-region rather than the entire image. However, our conference version does not address the optimal sub-region for hiding, which is crucial when dealing with local region concealment problems. In this paper extension, we consider human perceptual characteristics and introduce an optimal hiding region recommendation algorithm that comprehensively incorporates Just Noticeable Difference (JND) and visual saliency factors into consideration. Extensive experiments demonstrate superior visual quality and robustness compared to state-of-the-art methods. With the assistance of our proposed hiding region recommendation algorithm, concealed information becomes even less visible than the results of our conference version without compromising robustness.

Differences in forgery attributes of images generated in CNN-synthesized and image-editing domains are large, and such differences make a unified image forgery detection and localization (IFDL) challenging. To this end, we present a hierarchical fine-grained formulation for IFDL representation learning. Specifically, we first represent forgery attributes of a manipulated image with multiple labels at different levels. Then we perform fine-grained classification at these levels using the hierarchical dependency between them. As a result, the algorithm is encouraged to learn both comprehensive features and inherent hierarchical nature of different forgery attributes, thereby improving the IFDL representation. Our proposed IFDL framework contains three components: multi-branch feature extractor, localization and classification modules. Each branch of the feature extractor learns to classify forgery attributes at one level, while localization and classification modules segment the pixel-level forgery region and detect image-level forgery, respectively. Lastly, we construct a hierarchical fine-grained dataset to facilitate our study. We demonstrate the effectiveness of our method on 7 different benchmarks, for both tasks of IFDL and forgery attribute classification. Our source code and dataset can be found: github.com/CHELSEA234/HiFi-IFDL.

Recent deep learning-based optical flow estimators have exhibited impressive performance in generating local flows between consecutive frames. However, the estimation of long-range flows between distant frames, particularly under complex object deformation and large motion occlusion, remains a challenging task. One promising solution is to accumulate local flows explicitly or implicitly to obtain the desired long-range flow. Nevertheless, the accumulation errors and flow misalignment can hinder the effectiveness of this approach. This paper proposes a novel recurrent framework called AccFlow, which recursively backward accumulates local flows using a deformable module called as AccPlus. In addition, an adaptive blending module is designed along with AccPlus to alleviate the occlusion effect by backward accumulation and rectify the accumulation error. Notably, we demonstrate the superiority of backward accumulation over conventional forward accumulation, which to the best of our knowledge has not been explicitly established before. To train and evaluate the proposed AccFlow, we have constructed a large-scale high-quality dataset named CVO, which provides ground-truth optical flow labels between adjacent and distant frames. Extensive experiments validate the effectiveness of AccFlow in handling long-range optical flow estimation. Codes are available at https://github.com/mulns/AccFlow.

Recently, Users Generated Content (UGC) videos becomes ubiquitous in our daily lives. However, due to the limitations of photographic equipments and techniques, UGC videos often contain various degradations, in which one of the most visually unfavorable effects is the underexposure. Therefore, corresponding video enhancement algorithms such as Low-Light Video Enhancement (LLVE) have been proposed to deal with the specific degradation. However, different from video enhancement algorithms, almost all existing Video Quality Assessment (VQA) models are built generally rather than specifically, which measure the quality of a video from a comprehensive perspective. To the best of our knowledge, there is no VQA model specially designed for videos enhanced by LLVE algorithms. To this end, we first construct a Low-Light Video Enhancement Quality Assessment (LLVE-QA) dataset in which 254 original low-light videos are collected and then enhanced by leveraging 8 LLVE algorithms to obtain 2,060 videos in total. Moreover, we propose a quality assessment model specialized in LLVE, named Light-VQA. More concretely, since the brightness and noise have the most impact on low-light enhanced VQA, we handcraft corresponding features and integrate them with deep-learning-based semantic features as the overall spatial information. As for temporal information, in addition to deep-learning-based motion features, we also investigate the handcrafted brightness consistency among video frames, and the overall temporal information is their concatenation. Subsequently, spatial and temporal information is fused to obtain the quality-aware representation of a video. Extensive experimental results show that our Light-VQA achieves the best performance against the current State-Of-The-Art (SOTA) on LLVE-QA and public dataset. Dataset and Codes can be found at https://github.com/wenzhouyidu/Light-VQA.

Video shakiness is an unpleasant distortion of User Generated Content (UGC) videos, which is usually caused by the unstable hold of cameras. In recent years, many video stabilization algorithms have been proposed, yet no specific and accurate metric enables comprehensively evaluating the stability of videos. Indeed, most existing quality assessment models evaluate video quality as a whole without specifically taking the subjective experience of video stability into consideration. Therefore, these models cannot measure the video stability explicitly and precisely when severe shakes are present. In addition, there is no large-scale video database in public that includes various degrees of shaky videos with the corresponding subjective scores available, which hinders the development of Video Quality Assessment for Stability (VQA-S). To this end, we build a new database named StableDB that contains 1,952 diversely-shaky UGC videos, where each video has a Mean Opinion Score (MOS) on the degree of video stability rated by 34 subjects. Moreover, we elaborately design a novel VQA-S model named StableVQA, which consists of three feature extractors to acquire the optical flow, semantic, and blur features respectively, and a regression layer to predict the final stability score. Extensive experiments demonstrate that the StableVQA achieves a higher correlation with subjective opinions than the existing VQA-S models and generic VQA models. The database and codes are available at https://github.com/QMME/StableVQA.

Low-Light Video Enhancement (LLVE) has received considerable attention in recent years. One of the critical requirements of LLVE is inter-frame brightness consistency, which is essential for maintaining the temporal coherence of the enhanced video. However, most existing single-image-based methods fail to address this issue, resulting in flickering effect that degrades the overall quality after enhancement. Moreover, 3D Convolution Neural Network (CNN)-based methods, which are designed for video to maintain inter-frame consistency, are computationally expensive, making them impractical for real-time applications. To address these issues, we propose an efficient pipeline named FastLLVE that leverages the Look-Up-Table (LUT) technique to maintain inter-frame brightness consistency effectively. Specifically, we design a learnable Intensity-Aware LUT (IA-LUT) module for adaptive enhancement, which addresses the low-dynamic problem in low-light scenarios. This enables FastLLVE to perform low-latency and low-complexity enhancement operations while maintaining high-quality results. Experimental results on benchmark datasets demonstrate that our method achieves the State-Of-The-Art (SOTA) performance in terms of both image quality and inter-frame brightness consistency. More importantly, our FastLLVE can process 1,080p videos at 50+ Frames Per Second (FPS), which is 2 X faster than SOTA CNN-based methods in inference time, making it a promising solution for real-time applications. The code is available at https://github.com/Wenhao-Li-777/FastLLVE.

Adverse weather conditions such as haze can deteriorate the performance of autonomous driving and intelligent transport systems. As a potential remedy, we propose an enhanced multi-scale network, dubbed GridDehazeNet+, for single image dehazing. The proposed dehazing method does not rely on the Atmosphere Scattering Model (ASM), and an explanation as to why it is not necessarily performing the dimension reduction offered by this model is provided. GridDehazeNet+ consists of three modules: pre-processing, backbone, and post-processing. The trainable pre-processing module can generate learned inputs with better diversity and more pertinent features as compared to those derived inputs produced by hand-selected pre-processing methods. The backbone module implements multi-scale estimation with two major enhancements: 1) a novel grid structure that effectively alleviates the bottleneck issue via dense connections across different scales; 2) a spatial-channel attention block that can facilitate adaptive fusion by consolidating dehazing-relevant features. The post-processing module helps to reduce the artifacts in the final output. Due to domain shift, the model trained on synthetic data may not generalize well on real data. To address this issue, we shape the distribution of synthetic data to match that of real data, and use the resulting translated data to finetune our network. We also propose a novel intra-task knowledge transfer mechanism that can memorize and take advantage of synthetic domain knowledge to assist the learning process on the translated data. Experimental results demonstrate that the proposed method outperforms the state-of-the-art on several synthetic dehazing datasets, and achieves the superior performance on real-world hazy images after finetuning.

This paper presents a strategic approach to tackling trimap-free natural image matting. Specifically, to address the false detection issue of existing trimap-free matting algorithms when the foreground object is not uniquely defined, we design a novel tangled structure (TangleNet) to handle foreground detection and matting prediction simultaneously. TangleNet enables information exchange between foreground segmentation and alpha prediction, producing high-quality alpha mattes for the most salient foreground object based on RGB inputs alone. TangleNet boosts network performance with a frequency-guided attention mechanism utilizing wavelet data. Additionally, we pretrain for salient object detection to aid in the foreground segmentation. Experimental results demonstrate that TangleNet is on par with the state-of-the-art matting methods requiring additional inputs, and outperforms all previous trimap-free algorithms in terms of both qualitative and quantitative results.

Heart disease is a common condition worldwide and has become one of the leading causes of death worldwide. The electrocardiogram (PCG) is a safe, painless, and non-invasive test that captures bioacoustic information reflecting the function of the heart by capturing the acoustic signal of the patient’s heart. Nowadays, based on biosignal processing and artificial intelligence technologies, automated heart sound classification is playing an increasingly important role in clinical applications. In this paper, we propose a new parallel CNNs-transformer network with multi-scale feature context aggregation (PCTMF-Net). It combines the advantages of CNNs and transformer to achieve efficient heart sound classification. In PCTMF-Net, firstly, the heart tone signal features are extracted using the second-order spectral analysis, and a transformer-based MHTE-4 (multi-head transformer encoder with four attention heads) is designed to encode and aggregate the contextual information, and then, two CNNs feature extractors are designed in parallel with MHTE-4 to capture the hierarchical features. Finally, the feature vectors obtained from CNNs and MHTE-4 through feature fusion in PCTMF-Net will be fed into the fully connected layer for predicting the classification results of heart sounds. In addition, we perform validation based on two publicly available mutually exclusive heart sound datasets and conduct extensive experiments and comparisons of existing algorithms under different metrics. The experimental results show that our proposed method achieves 99.36% accuracy on the Yaseen dataset and 93% accuracy on the PhysioNet dataset. It surpasses current algorithms in terms of accuracy, recall and F 1-score metrics. The aim of this study is to apply these new techniques and methods to improve the diagnostic accuracy and validity of heart disease for clinical use.

Existing methods for video interpolation heavily rely on deep convolution neural networks, and thus suffer from their intrinsic limitations, such as content-agnostic kernel weights and restricted receptive field. To address these issues, we propose a Transformer-based video interpolation framework that allows content-aware aggregation weights and considers long-range dependencies with the self-attention operations. To avoid the high computational cost of global self-attention, we introduce the concept of local attention into video interpolation and extend it to the spatial-temporal domain. Furthermore, we propose a space-time separation strategy to save memory usage, which also improves performance. In addition, we develop a multi-scale frame synthesis scheme to fully realize the potential of Transformers. Extensive experiments demonstrate the proposed model performs favorably against the state-of-the-art methods both quantitatively and qualitatively on a variety of benchmark datasets. The code and models are released at https://github.com/zhshi0816/Video-Frame-Interpolation-Transformer.

Traditional biological and pharmaceutical manufacturing plants are controlled by human workers or pre-defined thresholds. Modernized factories have advanced process control algorithms such as model predictive control (MPC). However, there is little exploration of applying deep reinforcement learning to control manufacturing plants. One of the reasons is the lack of high fidelity simulations and standard APIs for benchmarking. To bridge this gap, we develop an easy-to-use library that includes five high-fidelity simulation environments: BeerFMTEnv, ReactorEnv, AtropineEnv, PenSimEnv and mAbEnv, which cover a wide range of manufacturing processes. We build these environments on published dynamics models. Furthermore, we benchmark online and offline, model-based and model-free reinforcement learning algorithms for comparisons of follow-up research.

To defend against manipulation of image content, such as splicing, copy-move, and removal, we develop a Progressive Spatio-Channel Correlation Network (PSCC-Net) to detect and localize image manipulations. PSCC-Net processes the image in a two-path procedure: a top-down path that extracts local and global features and a bottom-up path that detects whether the input image is manipulated, and estimates its manipulation masks at multiple scales, where each mask is conditioned on the previous one. Different from the conventional encoder-decoder and no-pooling structures, PSCC-Net leverages features at different scales with dense cross-connections to produce manipulation masks in a coarse-to-fine fashion. Moreover, a Spatio-Channel Correlation Module (SCCM) captures both spatial and channel-wise correlations in the bottom-up path, which endows features with holistic cues, enabling the network to cope with a wide range of manipulation attacks. Thanks to the light-weight backbone and progressive mechanism, PSCC-Net can process$1,080\text{P}$images at 50+FPS. Extensive experiments demonstrate the superiority of PSCC-Net over the state-of-the-art methods on both detection and localization. Codes and models are available athttps://github.com/proteus1991/PSCC-Net.

Video frame interpolation aims at synthesizing intermediate frames from nearby source frames while maintaining spatial and temporal consistencies. The existing deep-learning-based video frame interpolation methods can be roughly divided into two categories: flow-based methods and kernel-based methods. The performance of flow-based methods is often jeopardized by the inaccuracy of flow map estimation due to oversimplified motion models, while that of kernel-based methods tends to be constrained by the rigidity of kernel shape. To address these performance-limiting issues, a novel mechanism named generalized deformable convolution is proposed, which can effectively learn motion information in a data-driven manner and freely select sampling points in space-time. We further develop a new video frame interpolation method based on this mechanism. Our extensive experiments demonstrate that the new method performs favorably against the state-of-the-art, especially when dealing with complex motions. Code is available athttps://github.com/zhshi0816/GDConvNet.

To the best of our knowledge, the existing deep-learning-based Video Super-Resolution (VSR) methods exclusively make use of videos produced by the Image Signal Processor (ISP) of the camera system as inputs. Such methods are 1) inherently suboptimal due to information loss incurred by non-invertible operations in ISP, and 2) inconsistent with the real imaging pipeline where VSR in fact serves as a pre-processing unit of ISP. To address this issue, we propose a new VSR method that can directly exploit camera sensor data, accompanied by a carefully built Raw Video Dataset (RawVD) for training, validation, and testing. This method consists of a Successive Deep Inference (SDI) module and a reconstruction module, among others. The SDI module is designed according to the architectural principle suggested by a canonical decomposition result for Hidden Markov Model (HMM) inference; it estimates the target high-resolution frame by repeatedly performing pairwise feature fusion using deformable convolutions. The reconstruction module, built with elaborately designed Attention-based Residual Dense Blocks (ARDBs), serves the purpose of 1) refining the fused feature and 2) learning the color information needed to generate a spatial-specific transformation for accurate color correction. Extensive experiments demonstrate that owing to the informativeness of the camera raw data, the effectiveness of the network architecture, and the separation of super-resolution and color correction processes, the proposed method achieves superior VSR results compared to the state-of-the-art and can be adapted to any specific camera-ISP. Code and dataset are available at https://github.com/proteus1991/RawVSR.

Video super-resolution aims at generating a high-resolution video from its low-resolution counterpart. With the rapid rise of deep learning, many recently proposed video super-resolution methods use convolutional neural networks in conjunction with explicit motion compensation to capitalize on statistical dependencies within and across low-resolution frames. Two common issues of such methods are noteworthy. Firstly, the quality of the final reconstructed HR video is often very sensitive to the accuracy of motion estimation. Secondly, the warp grid needed for motion compensation, which is specified by the two flow maps delineating pixel displacements in horizontal and vertical directions, tends to introduce additional errors and jeopardize the temporal consistency across video frames. To address these issues, we propose a novel dynamic local filter network to perform implicit motion estimation and compensation by employing, via locally connected layers, sample-specific and position-specific dynamic local filters that are tailored to the target pixels. We also propose a global refinement network based on ResBlock and autoencoder structures to exploit non-local correlations and enhance the spatial consistency of super-resolved frames. The experimental results demonstrate that the proposed method outperforms the state-of-the-art, and validate its strength in terms of local transformation handling, temporal consistency as well as edge sharpness.

We propose an end-to-end trainable Convolutional Neural Network (CNN), named GridDehazeNet, for single image dehazing. The GridDehazeNet consists of three modules: pre-processing, backbone, and post-processing. The trainable pre-processing module can generate learned inputs with better diversity and more pertinent features as compared to those derived inputs produced by hand-selected pre-processing methods. The backbone module implements a novel attention-based multi-scale estimation on a grid network, which can effectively alleviate the bottleneck issue often encountered in the conventional multi-scale approach. The post-processing module helps to reduce the artifacts in the final output. Experimental results indicate that the GridDehazeNet outperforms the state-of-the-arts on both synthetic and real-world images. The proposed hazing method does not rely on the atmosphere scattering model, and we provide an explanation as to why it is not necessarily beneficial to take advantage of the dimension reduction offered by the atmosphere scattering model for image dehazing, even if only the dehazing results on synthetic images are concerned.

Multi-frame image super-resolution focuses on reconstructing a high-resolution image from a set of low-resolution images with high similarity. Combining image prior knowledge with fidelity model, the Bayesian-based methods have been considered as an effective technique in super-resolution. The minimization function derived from maximum a posteriori probability (MAP) is composed of a fidelity term and a regularization term. In this paper, based on the MAP estimation, we propose a novel initialization method for super-resolution imaging. For the fidelity term in our proposed method, the half-quadratic estimation is used to choose error norm adaptively instead of using fixed and norms. Besides, a spatial weight matrix is used as a confidence map to scale the estimation result. For the regularization term, we propose a novel regularization method based on adaptive bilateral total variation (ABTV). Both the fidelity term and the ABTV regularization guarantee the robustness of our framework. The fidelity term is mainly responsible for dealing with misregistration, blur, and other kinds of large errors, while the ABTV regularization aims at edge preservation and noise removal. The proposed scheme is tested on both synthetic data and real data. The experimental results illustrate the superiority of our proposed method in terms of edge preservation and noise removal over the state-of-the-art algorithms.