Scene regression methods, such as VGGT [86], solve the Structure-from-Motion (SfM) problem by directly regressing camera poses and 3D scene structures from input images. They demonstrate impressive performance in handling images under extreme viewpoint changes. However, these methods struggle to handle a large number of input images. To address this problem, we introduce SAIL-Recon, a feed-forward Transformer for large scale SfM, by augmenting the scene regression network with visual localization capabilities. Specifically, our method first computes a neural scene representation tokens from a subset of anchor images. The regression network is then fine-tuned to reconstruct all input images conditioned on this neural scene representation. Comprehensive experiments show that our method not only scales efficiently to large-scale scenes, but also achieves state-of-the-art results on both camera pose estimation and novel view synthesis benchmarks, including TUM-RGBD, CO3Dv2, and Tanks & Temples. Code and models are publicly available here.

Referring Remote Sensing Image Segmentation (RRSIS) aims to segment target objects in remote sensing (RS) images based on textual descriptions. Although Segment Anything Model 2 (SAM2) has shown remarkable performance in various segmentation tasks, its application to RRSIS presents several challenges, including understanding the text-described RS scenes and generating effective prompts from text. To address these issues, we propose RS2-SAM2, a novel framework that adapts SAM2 to RRSIS by aligning the adapted RS features and textual features while providing pseudo-mask-based dense prompts. Specifically, we employ a union encoder to jointly encode the visual and textual inputs, generating aligned visual and text embeddings as well as multimodal class tokens. A bidirectional hierarchical fusion module is introduced to adapt SAM2 to RS scenes and align adapted visual features with the visually enhanced text embeddings, improving the model's interpretation of text-described RS scenes. To provide precise target cues for SAM2, we design a mask prompt generator, which takes the visual embeddings and class tokens as input and produces a pseudo-mask as the dense prompt of SAM2. Experimental results on several RRSIS benchmarks demonstrate that RS2-SAM2 achieves state-of-the-art performance.

Recently, All-in-One image restoration (AiOIR) has advanced significantly, offering promising solutions for complex real-world degradations. However, most existing approaches heavily rely on degradation-specific representation learning, which can lead to oversmoothing and artifacts in the restored images. To address this limitation, we propose ClearAIR, a novel AiOIR framework inspired by human visual perception and designed with a hierarchical restoration strategy in a coarse-to-fine manner. First, leveraging the global priority characteristic of early human visual perception, we employ an image quality assessment model to evaluate the overall image structure and degradation level. Next, we introduce a Semantic Guidance Unit to provide coarse semantic region guidance and a Task Identifier to predict local degradation types, enabling a more informed characterization of local degradation patterns. Finally, aiming at the challenge of local detail restoration, we propose an Internal Clue Reuse Mechanism that deeply mines the internal information of the image in a self-supervised manner to enhance the model’s capacity for fine-detail recovery. Experimental results demonstrate that ClearAIR achieves superior restoration performance across diverse synthetic and real-world datasets.

Existing All-in-One image restoration methods often fail to perceive degradation types and severity levels simultaneously, overlooking the importance of fine-grained quality perception. Moreover, these methods often utilize highly customized backbones, which hinder their adaptability and integration into more advanced restoration networks. To address these limitations, we propose Perceive-IR, a novel backbone-agnostic All-in-One image restoration framework designed for fine-grained quality control across various degradation types and severity levels. Its modular structure allows core components to function independently of specific backbones, enabling seamless integration into advanced restoration models without significant modifications. Specifically, Perceive-IR operates in two key stages: 1) multi-level quality-driven prompt learning stage, where a fine-grained quality perceiver is meticulously trained to discern three-tier quality levels by optimizing the alignment between prompts and images within the CLIP perception space. This stage ensures a nuanced understanding of image quality, laying the groundwork for subsequent restoration; 2) restoration stage, where the quality perceiver is seamlessly integrated with a difficulty-adaptive perceptual loss, forming a quality-aware learning strategy. This strategy not only dynamically differentiates sample learning difficulty but also achieves fine-grained quality control by driving the restored image toward the ground truth while pulling it away from both low- and medium-quality samples. Furthermore, Perceive-IR incorporates a Semantic Guidance Module (SGM) and Compact Feature Extraction (CFE). The SGM leverages semantic information from pre-trained vision models to provide high-level contextual guidance, while the CFE focuses on extracting degradation-specific features, ensuring accurate handling of diverse image degradations. Extensive experiments demonstrate that Perceive-IR not only surpasses state-of-the-art methods but also generalizes reliably to zero-shot real-world and unknown degraded scenes, while adapting seamlessly to different backbone networks. This versatility underscores the framework's robustness and backbone-agnostic design. Project page at https://house-yuyu.github.io/Perceive-IR/.

Recently, the diffusion model has emerged as a powerful generative technique for robotic policy learning, capable of modeling multi-mode action distributions. Leveraging its capability for end-to-end autonomous driving is a promising direction. However, the numerous denoising steps in the robotic diffusion policy and the more dynamic, open-world nature of traffic scenes pose substantial challenges for generating diverse driving actions at a real-time speed. To address these challenges, we propose a novel truncated diffusion policy that incorporates prior multi-mode anchors and truncates the diffusion schedule, enabling the model to learn denoising from anchored Gaussian distribution to the multi-mode driving action distribution. Additionally, we design an efficient cascade diffusion decoder for enhanced interaction with conditional scene context. The proposed model, DiffusionDrive, demonstrates 10× reduction in denoising steps compared to vanilla diffusion policy, delivering superior diversity and quality in just 2 steps. On the planning-oriented NAVSIM dataset, with aligned ResNet-34 backbone, DiffusionDrive achieves 88.1 PDMS without bells and whistles, setting a new record, while running at a real-time speed of 45 FPS on an NVIDIA 4090. Qualitative results on challenging scenarios further confirm that DiffusionDrive can robustly generate diverse plausible driving actions.

Referring video object segmentation (RVOS) aims to segment objects in a video according to textual descriptions, which requires the integration of multimodal information and temporal dynamics perception. The Segment Anything Model 2 (SAM 2) has shown great effectiveness across various video segmentation tasks. However, its application to offline RVOS is challenged by the translation of the text into effective prompts and a lack of global context awareness. In this paper, we propose a novel RVOS framework, termed MPG-SAM 2, to address these challenges. Specifically, MPG-SAM 2 employs a unified multimodal encoder to jointly encode video and textual features, generating semantically aligned video and text embeddings, along with multimodal class tokens. A mask prior generator utilizes the video embeddings and class tokens to create pseudo masks of target objects and global context. These masks are fed into the prompt encoder as dense prompts along with multimodal class tokens as sparse prompts to generate accurate prompts for SAM 2. To provide the online SAM 2 with a global view, we introduce a hierarchical global-historical aggregator, which allows SAM 2 to aggregate global and historical information of target objects at both pixel and object levels, enhancing the target representation and temporal consistency. Extensive experiments on several RVOS benchmarks demonstrate the superiority of MPG-SAM 2 and the effectiveness of our proposed modules. The code is available at https://github.com/rongfu-dsb/MPG-SAM2.

In the field of autonomous driving, sensor simulation is essential for generating rare and diverse scenarios that are difficult to capture in real-world environments. Current solutions fall into two categories: 1) CG-based methods, such as CARLA, which lack diversity and struggle to scale to the vast array of rare cases required for robust perception training; and 2) learning-based approaches, such as NeuSim, which are limited to specific object categories (vehicles) and require extensive multi-sensor data, hindering their applicability to generic objects. To address these limitations, we propose SynthDrive, a scalable "real2sim2real" system that leverages 3D generation to automate asset mining, generation, and rare-case data synthesis.Our framework introduces two key innovations: 1) Automated Rare-Case Mining and Synthesis. Given a text prompt describing specific objects, SynthDrive automatically mines image data from the Internet and then generates corresponding high-fidelity 3D assets, which eliminates the need for costly manual data collection. By integrating these assets into existing street-view data, our pipeline produces photorealistic rare-case data, supporting rapid scaling to diverse assets including irregular obstacles and temporary traffic facilities. 2) High-Fidelity 3D Generation. We propose a hybrid asset generation pipeline that combines a geometry-aware LRM, iterative mesh optimization, and an improved texture fusion algorithm. Our approach achieves 0.0164 Chamfer Distance on the GSO dataset, outperforming InstantMesh by 14.1% in geometry accuracy, and achieves 19.05 PSNR (vs 16.84) for texture quality. This enables fine geometry details and high-resolution texture generation, which is essential for perception model training. Experiments demonstrate that SynthDrive-generated data improves the performance of downstream perception tasks (2D and 3D detection on rare objects) by 2-4% mAP. SynthDrive greatly lowers the data production cost and improves the diversity for corner-case data generation, showcasing extensive potential applications in the field of autonomous driving.

Existing underwater image restoration (UIR) methods generally only handle color distortion or jointly address color and haze issues, but they often overlook the more complex degradations that can occur in underwater scenes. To address this limitation, we propose a Universal Underwater Image Restoration method, termed as UniUIR, considering the complex scenario of real-world underwater mixed distortions as an all-in-one manner. To disentangle degradation-specific effects and capture their inter-correlations, we propose the Mamba Mixture-of-Experts module (MMoEM). Each expert specializes in distinct aspects of degradation, while gating mechanism dynamically routes features to appropriate experts. This design enables collaborative prior extraction and preserves global context, all within linear computational complexity. Building upon this foundation, to enhance degradation representation and address the task conflicts that arise when handling multiple types of degradation, we introduce the spatial-frequency prior generator. This module extracts degradation prior information in both spatial and frequency domains, and adaptively selects the most appropriate task-specific prompts based on image content, thereby improving the accuracy of image restoration. Finally, to more effectively address complex, region-dependent distortions in UIR task, we incorporate depth information derived from a large-scale pre-trained depth prediction model, thereby enabling the network to perceive and leverage depth variations across different image regions to handle localized degradation. Extensive experiments demonstrate that UniUIR can produce more attractive results across qualitative and quantitative comparisons, and shows strong generalization than state-of-the-art methods. Project page at https://house-yuyu.github.io/UniUIR.

RAW to sRGB mapping, which aims to convert RAW images from smartphones into RGB form equivalent to that of Digital Single-Lens Reflex (DSLR) cameras, has become an important area of research. However, current methods often ignore the difference between cell phone RAW images and DSLR camera RGB images, a difference that goes beyond the color matrix and extends to spatial structure due to resolution variations. Recent methods directly rebuild color mapping and spatial structure via shared deep representation, limiting optimal performance. Inspired by Image Signal Processing (ISP) pipeline, which distinguishes image restoration and enhancement, we present a novel Neural ISP framework, named FourierISP. This approach breaks the image down into style and structure within the frequency domain, allowing for independent optimization. FourierISP is comprised of three subnetworks: Phase Enhance Subnet for structural refinement, Amplitude Refine Subnet for color learning, and Color Adaptation Subnet for blending them in a smooth manner. This approach sharpens both color and structure, and extensive evaluations across varied datasets confirm that our approach realizes state-of-the-art results. Code will be available at https://github.com/alexhe101/FourierISP.

Recently the state space models (SSMs) with efficient hardware-aware designs, i.e., the Mamba deep learning model, have shown great potential for long sequence modeling. Meanwhile building efficient and generic vision backbones purely upon SSMs is an appealing direction. However, representing visual data is challenging for SSMs due to the position-sensitivity of visual data and the requirement of global context for visual understanding. In this paper, we show that the reliance on self-attention for visual representation learning is not necessary and propose a new generic vision backbone with bidirectional Mamba blocks (Vim), which marks the image sequences with position embeddings and compresses the visual representation with bidirectional state space models. On ImageNet classification, COCO object detection, and ADE20k semantic segmentation tasks, Vim achieves higher performance compared to well-established vision transformers like DeiT, while also demonstrating significantly improved computation & memory efficiency. For example, Vim is 2.8x faster than DeiT and saves 86.8% GPU memory when performing batch inference to extract features on images with a resolution of 1248x1248. The results demonstrate that Vim is capable of overcoming the computation & memory constraints on performing Transformer-style understanding for high-resolution images and it has great potential to be the next-generation backbone for vision foundation models.

The recent development of online static map element (a.k.a. HD Map) construction algorithms has raised a vast demand for data with ground truth annotations. However, available public datasets currently cannot provide high-quality training data regarding consistency and accuracy. To this end, we present CAMA: a vision-centric approach for Consistent and Accurate Map Annotation. Without LiDAR inputs, our proposed framework can still generate high-quality 3D annotations of static map elements. Specifically, the annotation can achieve high reprojection accuracy across all surrounding cameras and is spatial-temporal consistent across the whole sequence. We apply our proposed framework to the popular nuScenes dataset to provide efficient and highly accurate annotations. Compared with the original nuScenes static map element, models trained with annotations from CAMA achieve lower reprojection errors (e.g., 4.73 vs. 8.03 pixels).

Driving scene topology reasoning aims to understand the objects present in the current road scene and model their topology relationships to provide guidance information for downstream tasks. Previous approaches fail to adequately facilitate interactions among traffic objects and neglect to incorporate scene information into topology reasoning, thus limiting the comprehensive exploration of potential correlations among objects and diminishing the practical significance of the reasoning results. Besides, the lack of constraints on lane direction may introduce erroneous guidance information and lead to a decrease in topology prediction accuracy. In this paper, we propose a novel topology reasoning framework, dubbed TSTGT, to address these issues. Specifically, we design a divide-and-conquer topology graph Transformer to respectively infer the lane-lane and lane-traffic topology relationships, which can effectively aggregate the local and global object information in the driving scene and facilitate the topology relationship learning. Additionally, a traffic scene-assisted reasoning module is devised and combined with the topology graph Transformer to enhance the practical significance of lane-traffic topology. In terms of lane detection, we develop a point-wise matching strategy to infer lane centerlines with correct directions, thereby improving the topology reasoning accuracy. Extensive experimental results on Openlane-V2 benchmark validate the superiority of our TSTGT over state-of-the-art methods and the effectiveness of our proposed modules. The code is available at https://github.com/rongfu-dsb/TSTGT.

Labeling objects with pixel-wise segmentation requires a huge amount of human labor compared to bounding boxes. Most existing methods for weakly supervised instance segmentation focus on designing heuristic losses with priors from bounding boxes. While, we find that box-supervised methods can produce some fine segmentation masks and we wonder whether the detectors could learn from these fine masks while ignoring low-quality masks. To answer this question, we present BoxTeacher, an efficient and end-to-end training framework for high-performance weakly supervised instance segmentation, which leverages a sophisticated teacher to generate high-quality masks as pseudo labels. Considering the massive noisy masks hurt the training, we present a mask-aware confidence score to estimate the quality of pseudo masks, and propose the noiseaware pixel loss and noise-reduced affinity loss to adaptively optimize the student with pseudo masks. Extensive experiments can demonstrate effectiveness of the proposed BoxTeacher. Without bells and whistles, BoxTeacher remarkably achieves 35.0 mask AP and 36.5 mask AP with ResNet-50 and ResNet-101 respectively on the challenging COCO dataset, which outperforms the previous state-of-the-art methods by a significant margin and bridges the gap between box-supervised and mask-supervised methods.

Bird's Eye View (BEV) semantic segmentation is a critical task in autonomous driving. However, existing Transformer-based methods confront difficulties in transforming Perspective View (PV) to BEV due to their unidirectional and posterior interaction mechanisms. To address this issue, we propose a novel Bi-directional and Early Interaction Transformers framework named BAEFormer, consisting of (i) an early-interaction PV-BEV pipeline and (ii) a bi-directional cross-attention mechanism. Moreover, we find that the image feature maps' resolution in the cross-attention module has a limited effect on the final performance. Under this critical observation, we propose to enlarge the size of input images and downsample the multi-view image features for cross-interaction, further improving the accuracy while keeping the amount of computation controllable. Our proposed method for BEV semantic segmentation achieves state-of-the-art performance in real-time inference speed on the nuScenes dataset, i.e., 38.9 mIoU at 45 FPS on a single A100 GPU.

Autonomous driving requires a comprehensive understanding of the surrounding environment for reliable trajectory planning. Previous works rely on dense rasterized scene representation (e.g., agent occupancy and semantic map) to perform planning, which is computationally intensive and misses the instance-level structure information. In this paper, we propose VAD, an end-to-end vectorized paradigm for autonomous driving, which models the driving scene as a fully vectorized representation. The proposed vectorized paradigm has two significant advantages. On one hand, VAD exploits the vectorized agent motion and map elements as explicit instance-level planning constraints which effectively improves planning safety. On the other hand, VAD runs much faster than previous end-to-end planning methods by getting rid of computation-intensive rasterized representation and hand-designed post-processing steps. VAD achieves state-of-the-art end-to-end planning performance on the nuScenes dataset, outperforming the previous best method by a large margin. Our base model, VAD-Base, greatly reduces the average collision rate by 29.0% and runs 2.5× faster. Besides, a lightweight variant, VAD-Tiny, greatly improves the inference speed (up to 9.3×) while achieving comparable planning performance. We believe the excellent performance and the high efficiency of VAD are critical for the real-world deployment of an autonomous driving system. Code and models are available at https://github.com/hustvl/VAD for facilitating future research.

Response-based Knowledge Distillation refers to the technique of supervising the student network with the teacher networks' predictions. The method is motivated by observing that the predicted probabilities reflect the relation among labels, which is the knowledge to be transferred. This paper explores the transferred knowledge from a novel perspective: comparing the knowledge transferred through different teachers. Two intriguing properties are observed. First, higher confidence scores of teachers' predictions lead to better distillation results, and second, teachers' incorrectly predicted training samples should be kept for distillation. We then analyze the phenomenon by studying teachers' decision boundaries, of which some can help the student generalize while some may not. Based on the observations, we further propose an embarrassingly simple distillation framework named Efficient Distillation, which is effective on ImageNet with different teacher-student pairs: When using ResNet34 as the teacher, the student ResNet18 trained from scratch reaches 74.07% Top-1 accuracy within 98 GPU hours (RTX 3090), outperforming current state-of-the-art result (73.19%) by a large margin. Our code is available at https://github.com/lsongx/EffDstl.

As the size of circuit designs continues to grow rapidly, artificial intelligence technologies are being extensively used in Electronic Design Automation (EDA) to assist with circuit design. Placement and routing are the most time-consuming parts of the physical design process, and how to quickly evaluate the placement has become a hot research topic. Prior works either transformed circuit designs into images using hand-crafted methods and then used Convolutional Neural Networks (CNN) to extract features, which are limited by the quality of the hand-crafted methods and could not achieve end-to-end training, or treated the circuit design as a graph structure and used Graph Neural Networks (GNN) to extract features, which require time-consuming preprocessing. In our work, we propose a novel perspective for circuit design by treating circuit components as point clouds and using Transformer-based point cloud perception methods to extract features from the circuit. This approach enables direct feature extraction from raw data without any preprocessing, allows for end-to-end training, and results in high performance. Experimental results show that our method achieves state-of-the-art performance in congestion prediction tasks on both the CircuitNet and ISPD2015 datasets, as well as in design rule check (DRC) violation prediction tasks on the CircuitNet dataset. Our method establishes a bridge between the relatively mature point cloud perception methods and the fast-developing EDA algorithms, enabling us to leverage more collective intelligence to solve this task. To facilitate the research of open EDA design, source codes and pre-trained models are released at https://github.com/hustvl/circuitformer.

The teacher-free online Knowledge Distillation (KD) aims to train an ensemble of multiple student models collaboratively and distill knowledge from each other. Although existing online KD methods achieve desirable performance, they often focus on class probabilities as the core knowledge type, ignoring the valuable feature representational information. We present a Mutual Contrastive Learning (MCL) framework for online KD. The core idea of MCL is to perform mutual interaction and transfer of contrastive distributions among a cohort of networks in an online manner. Our MCL can aggregate cross-network embedding information and maximize the lower bound to the mutual information between two networks. This enables each network to learn extra contrastive knowledge from others, leading to better feature representations, thus improving the performance of visual recognition tasks. Beyond the final layer, we extend MCL to intermediate layers and perform an adaptive layer-matching mechanism trained by meta-optimization. Experiments on image classification and transfer learning to visual recognition tasks show that layer-wise MCL can lead to consistent performance gains against state-of-the-art online KD approaches. The superiority demonstrates that layer-wise MCL can guide the network to generate better feature representations. Our code is publicly avaliable at https://github.com/winycg/L-MCL.

Estimating the 3D structure of the drivable surface and surrounding environment is a crucial task for assisted and autonomous driving. It is commonly solved either by using 3D sensors such as LiDAR or directly predicting the depth of points via deep learning. However, the former is expensive, and the latter lacks the use of geometry information for the scene. In this paper, instead of following existing methodologies, we propose Road Planar Parallax Attention Network (RPANet), a new deep neural network for 3D sensing from monocular image sequences based on planar parallax, which takes full advantage of the omnipresent road plane geometry in driving scenes. RPANet takes a pair of images aligned by the homography of the road plane as input and outputs a γ map (the ratio of height to depth) for 3D reconstruction. The γ map has the potential to construct a two-dimensional transformation between two consecutive frames. It implies planar parallax and can be combined with the road plane serving as a reference to estimate the 3D structure by warping the consecutive frames. Furthermore, we introduce a novel cross-attention module to make the network better perceive the displacements caused by planar parallax. To verify the effectiveness of our method, we sample data from the Waymo Open Dataset and construct annotations related to planar parallax. Comprehensive experiments are conducted on the sampled dataset to demonstrate the 3D reconstruction accuracy of our approach in challenging scenarios.

Due to the scarcity of annotated samples, the diversity between support set and query set becomes the main obstacle for few shot semantic segmentation. Most existing prototype-based approaches only exploit the prototype from the support feature and ignore the information from the query sample, failing to remove this obstacle.In this paper, we proposes a dual prototype network (DPNet) to dispose of few shot semantic segmentation from a new perspective. Along with the prototype extracted from the support set, we propose to build the pseudo-prototype based on foreground features in the query image. To achieve this goal, the cycle comparison module is developed to select reliable foreground features and generate the pseudo-prototype with them. Then, a prototype interaction module is utilized to integrate the information of the prototype and the pseudo-prototype based on their underlying correlation. Finally, a multi-scale fusion module is introduced to capture contextual information during the dense comparison between prototype (pseudo-prototype) and query feature. Extensive experiments conducted on two benchmarks demonstrate that our method exceeds previous state-of-the-arts with a sizable margin, verifying the effectiveness of the proposed method.

Multi-person pose estimation methods generally follow top-down and bottom-up paradigms, both of which can be considered as two-stage approaches thus leading to the high computation cost and low efficiency. Towards a compact and efficient pipeline for multi-person pose estimation task, in this paper, we propose to represent the human parts as points and present a novel body representation, which leverages an adaptive point set including the human center and seven human-part related points to represent the human instance in a more fine-grained manner. The novel representation is more capable of capturing the various pose deformation and adaptively factorizes the long-range center-to-joint displacement thus delivers a single-stage differentiable network to more precisely regress multi-person pose, termed as AdaptivePose. For inference, our proposed network eliminates the grouping as well as refinements and only needs a single-step disentangling process to form multi-person pose. Without any bells and whistles, we achieve the best speed-accuracy trade-offs of 67.4% AP / 29.4 fps with DLA-34 and 71.3% AP / 9.1 fps with HRNet-W48 on COCO test-dev dataset.

Off-the-shelf single-stage multi-person pose regression methods generally leverage the instance score (i.e., confidence of the instance localization) to indicate the pose quality for selecting the pose candidates. We consider that there are two gaps involved in existing paradigm: 1) The instance score is not well interrelated with the pose regression quality. 2) The instance feature representation, which is used for predicting the instance score, does not explicitly encode the structural pose information to predict the reasonable score that represents pose regression quality. To address the aforementioned issues, we propose to learn the pose regression quality-aware representation. Concretely, for the first gap, instead of using the previous instance confidence label (e.g., discrete {1,0} or Gaussian representation) to denote the position and confidence for person instance, we firstly introduce the Consistent Instance Representation (CIR) that unifies the pose regression quality score of instance and the confidence of background into a pixel-wise score map to calibrates the inconsistency between instance score and pose regression quality. To fill the second gap, we further present the Query Encoding Module (QEM) including the Keypoint Query Encoding (KQE) to encode the positional and semantic information for each keypoint and the Pose Query Encoding (PQE) which explicitly encodes the predicted structural pose information to better fit the Consistent Instance Representation (CIR). By using the proposed components, we significantly alleviate the above gaps. Our method outperforms previous single-stage regression-based even bottom-up methods and achieves the state-of-the-art result of 71.7 AP on MS COCO test-dev set.

Studying the inherent symmetry of data is of great importance in machine learning. Point cloud, the most important data format for 3D environmental perception, is naturally endowed with strong radial symmetry. In this work, we exploit this radial symmetry via a divide-and-conquer strategy to boost 3D perception performance and ease optimization. We propose Azimuth Normalization (AziNorm), which normalizes the point clouds along the radial direction and eliminates the variability brought by the difference of azimuth. AziNorm can be flexibly incorporated into most LiDAR-based perception methods. To validate its effectiveness and generalization ability, we apply AziNorm in both object detection and semantic segmentation. For detection, we integrate AziNorm into two representative detection methods, the one-stage SECOND detector and the state-of-the-art two-stage PV-RCNN detector. Experiments on Waymo Open Dataset demonstrate that AziNorm improves SECOND and PV-RCNN by 7.03 mAPH and 3.01 mAPH respectively. For segmentation, we integrate AziNorm into KPConv. On SemanticKitti dataset, AziNorm improves KPConv by 1.6/1.1 mIoU on val/test set. Besides, AziNorm remarkably improves data efficiency and accelerates convergence, reducing the requirement of data amounts or training epochs by an order of magnitude. SECOND w/ AziNorm can significantly outperform fully trained vanilla SECOND, even trained with only 10% data or 10% epochs. Code and models are available at https://github.com/hustvl/AziNorm.

In this paper, we propose a conceptually novel, efficient, and fully convolutional framework for real-time instance segmentation. Previously, most instance segmentation methods heavily rely on object detection and perform mask prediction based on bounding boxes or dense centers. In contrast, we propose a sparse set of instance activation maps, as a new object representation, to high-light informative regions for each foreground object. Then instance-level features are obtained by aggregating features according to the highlighted regions for recognition and segmentation. Moreover, based on bipartite matching, the instance activation maps can predict objects in a one-to-one style, thus avoiding non-maximum suppression (NMS) in post-processing. Owing to the simple yet effective designs with instance activation maps, SparseInst has extremely fast inference speed and achieves 40 FPS and 37.9 AP on the COCO benchmark, which significantly out-performs the counterparts in terms of speed and accuracy. Code and models are available at https://github.com/hustvl/SparseInst.

The performance of nighttime semantic segmentation is restricted by the poor illumination and a lack of pixel-wise annotation, which severely limit its application in autonomous driving. Existing works, e.g., using the twilight as the intermediate target domain to perform the adaptation from daytime to nighttime, may fail to cope with the inherent difference between datasets caused by the camera equipment and the urban style. Faced with these two types of domain shifts, i.e., the illumination and the inherent difference of the datasets, we propose a novel domain adaptation framework via cross-domain correlation distillation, called CCDistill. The invariance of illumination or inherent difference between two images is fully explored so as to make up for the lack of labels for nighttime images. Specifically, we extract the content and style knowledge contained in features, calculate the degree of inherent or illumination difference between two images. The domain adaptation is achieved using the invariance of the same kind of difference. Extensive experiments on Dark Zurich and ACDC demon-strate that CCDistill achieves the state-of-the-art performance for nighttime semantic segmentation. Notably, our method is a one-stage domain adaptation network which can avoid affecting the inference time. Our implementation is available at https://github.com/ghuan99/CCDistill.

Current Knowledge Distillation (KD) methods for semantic segmentation often guide the student to mimic the teacher's structured information generated from individual data samples. However, they ignore the global semantic relations among pixels across various images that are valuable for KD. This paper proposes a novel Cross-Image Relational KD (CIRKD), which focuses on transferring structured pixel-to-pixel and pixel-to-region relations among the whole images. The motivation is that a good teacher network could construct a well-structured feature space in terms of global pixel dependencies. CIRKD makes the student mimic better structured semantic relations from the teacher, thus improving the segmentation performance. Experimental results over Cityscapes, CamVid and Pascal VOC datasets demonstrate the effectiveness of our proposed approach against state-of-the-art distillation methods. The code is available at https://github.com/winycg/CIRKD.

In order to get raw images of high quality for downstream Image Signal Process (ISP), in this paper we present an Efficient Locally Multiplicative Transformer called ELMformer for raw image restoration. ELMformer contains two core designs especially for raw images whose primitive attribute is single-channel. The first design is a Bi-directional Fusion Projection (BFP) module, where we consider both the color characteristics of raw images and spatial structure of single-channel. The second one is that we propose a Locally Multiplicative Self-Attention (L-MSA) scheme to effectively deliver information from the local space to relevant parts. ELMformer can efficiently reduce the computational consumption and perform well on raw image restoration tasks. Enhanced by these two core designs, ELMformer achieves the highest performance and keeps the lowest FLOPs on raw denoising and raw deblurring benchmarks compared with state-of-the-arts. Extensive experiments demonstrate the superiority and generalization ability of ELMformer. On SIDD benchmark, our method has even better denoising performance than ISP-based methods which need huge amount of additional sRGB training images.

Object detection and instance segmentation with a large number of object categories and long-tailed data distribution are challenging for most existing deep learning models. As the number of classes increases, the outputs of a classifier become sensitive to likely noisy logits, which can easily result in an incorrect recognition. To alleviate the large-vocabulary problem, we cluster fine-grained classes into coarser parent classes and then build a classification tree to classify an object into a fine-grained class via its parent class. Because the number of parent class is much fewer, their logits are more stable to suppress the wrong/noisy logits existed in the fine-grained class nodes. Due to a variety of ways for clustering fine-grained classes into parent classes, we can further construct multiple trees to build a classification forest where each single tree contributes its vote to the fine-grained classification. Moreover, a simple yet effective resampling method, termed as NMS Resampling, is proposed aiming at solving the long tail (data imbalance) problem. Our method, coined as ForestDet, serves as a plug-and-play module, which can be readily employed in both one-stage and two-stage object recognition models for recognizing more than 1000 categories. Extensive experiments are conducted on the large vocabulary dataset LVIS. Compared to the Mask R-CNN baseline, our two-stage counterpart Forest R-CNN significantly boosts the performance by 11.5% and 3.9% AP improvements on the rare categories and overall categories, respectively. Compared to the RetinaNet baseline, our one-stage counterpart Forest RetinaNet improves 2.1% AP on overall categories. Moreover, we achieve state-of-the-art results on the LVIS dataset.Code and models are available athttps://github.com/JialianW/Forest_RCNN.

The problem of unsupervised domain adaptation in semantic segmentation is a major challenge for numerous computer vision tasks because acquiring pixel-level labels is time-consuming with expensive human labor. A large gap exists among data distributions in different domains, which will cause severe performance loss when a model trained with synthetic data is generalized to real data. Hence, we propose a novel domain adaptation approach, called Content Invariant Representation Network, to narrow the domain gap between the source (S) and target (T) domains. The previous works developed a network to directly transfer the knowledge from the S to T. On the contrary, the proposed method aims to progressively reduce the gap between S and T on the basis of a Content Invariant Representation (CIR). CIR is an intermediate domain (I) sharing invariant content with S and having similar data distribution to T. Then, an Ancillary Classifier Module (ACM) is designed to focus on pixel-level details and generate attention-aware results. ACM adaptively assigns different weights to pixels according to their domain offsets, thereby reducing local domain gaps. The global domain gap between CIR and T is also narrowed by enforcing local alignments. Last, we perform self-supervised training in the pseudo-labeled target domain to further fit the distribution of the real data. Comprehensive experiments on two domain adaptation tasks, that is, GTAV → Cityscapes and SYNTHIA → Cityscapes, clearly demonstrate the superiority of our method compared with state-of-the-art methods.

When adopting deep neural networks for a new vision task, a common practice is to start with fine-tuning some off-the-shelf well-trained network models from the community. Since a new task may require training a different network architecture with new domain data, taking advantage of off-the-shelf models is not trivial and generally requires considerable try-and-error and parameter tuning. In this paper, we denote a well-trained model as a teacher network and a model for the new task as a student network. We aim to ease the efforts of transferring knowledge from the teacher to the student network, robust to the gaps between their network architectures, domain data, and task definitions. Specifically, we propose a hybrid forward scheme in training the teacher-student models, alternately updating layer weights of the student model. The key merit of our hybrid forward scheme is on the dynamical balance between the knowledge transfer loss and task specific loss in training. We demonstrate the effectiveness of our method on a variety of tasks, e.g., model compression, segmentation, and detection, under a variety of knowledge transfer settings.

Face recognition has achieved a great success in recent years, it is still challenging to recognize those facial images with extreme poses. Traditional methods consider it as a domain gap problem. Many of them settle it by generating fake frontal faces from extreme ones, whereas they are tough to maintain the identity information with high computational consumption and uncontrolled disturbances. Our experimental analysis shows a dramatic precision drop with extreme poses. Meanwhile, those extreme poses just exist minor visual differences after small rotations. Derived from this insight, we attempt to relieve such a huge precision drop by making minor changes to the input images without modifying existing discriminators. A novel lightweight pseudo facial generation is proposed to relieve the problem of extreme poses without generating any frontal facial image. It can depict the facial contour information and make appropriate modifications to preserve the critical identity information. Specifically, the proposed method reconstructs pseudo profile faces by minimizing the pixel-wise differences with original profile faces and maintaining the identity consistent information from their corresponding frontal faces simultaneously. The proposed framework can improve existing discriminators and obtain a great promotion on several benchmark datasets.

Instance segmentation on point clouds is a fundamental task in 3D scene perception. In this work, we propose a concise clustering-based framework named HAIS, which makes full use of spatial relation of points and point sets. Considering clustering-based methods may result in over-segmentation or under-segmentation, we introduce the hierarchical aggregation to progressively generate instance proposals, i.e., point aggregation for preliminarily clustering points to sets and set aggregation for generating complete instances from sets. Once the complete 3D instances are obtained, a sub-network of intra-instance prediction is adopted for noisy points filtering and mask quality scoring. HAIS is fast (only 410ms per frame on Titan X)) and does not require non-maximum suppression. It ranks 1st on the ScanNet v2 benchmark1, achieving the highest 69.9% AP50and surpassing previous state-of-the-art (SOTA) methods by a large margin. Besides, the SOTA results on the S3DIS dataset validate the good generalization ability. Code is available at https://github.com/hustvl/HAIS.

Multi-view pedestrian detection aims to predict a bird’s eye view (BEV) occupancy map from multiple camera views. This task is confronted with two challenges: how to establish the 3D correspondences from views to the BEV map and how to assemble occupancy information across views. In this paper, we propose a novel Stacked HOmography Transformations (SHOT) approach, which is motivated by approximating projections in 3D world coordinates via a stack of homographies. We first construct a stack of transformations for projecting views to the ground plane at different height levels. Then we design a soft selection module so that the network learns to predict the likelihood of the stack of transformations. Moreover, we provide an in-depth theoretical analysis on constructing SHOT and how well SHOT approximates projections in 3D world coordinates. SHOT is empirically verified to be capable of estimating accurate correspondences from individual views to the BEV map, leading to new state-of-the-art performance on standard evaluation benchmarks.

In this work, we propose a novel deep online correction (DOC) framework for monocular visual odometry. The whole pipeline has two stages: First, depth maps and initial poses are obtained from convolutional neural networks (CNNs) trained in self-supervised manners. Second, the poses predicted by CNNs are further improved by minimizing photometric errors via gradient updates of poses during inference phases. The benefits of our proposed method are twofold: 1) Different from online-learning methods, DOC does not need to calculate gradient propagation for parameters of CNNs. Thus, it saves more computation resources during inference phases. 2) Unlike hybrid methods that combine CNNs with traditional methods, DOC fully relies on deep learning (DL) frameworks. Though without complex back-end optimization modules, our method achieves outstanding performance with relative transform error (RTE) = 2.0% on KITTI Odometry benchmark for Seq. 09, which outperforms traditional monocular VO frameworks and is comparable to hybrid methods.

Multi-label image classification aims to predict multiple labels for a single image. However, the difficulties of predicting different labels may vary dramatically due to semantic variations of the label as well as the image context. Direct learning of multi-label classification models has the risk of being biased and overfitting those difficult labels, e.g., deep network based classifiers are over-trained on the difficult labels, therefore, lead to false-positive errors of those difficult labels during testing. To handle difficult labels of multi-label image classification, we propose to calibrate the model, which not only predicts the labels but also estimates the uncertainty of the prediction. With the new calibration branch of the network, the classification model is trained with the pick-all-labels normalized loss and optimized pertaining to the number of positive labels. Moreover, to improve performance on difficult labels, instead of annotating them, we leverage the calibrated model as the teacher network and teach the student network about handling difficult labels via uncertainty distillation. Our proposed uncertainty distillation teaches the student network which labels are highly uncertain through prediction distribution distillation, and locates the image regions that cause such uncertain predictions through uncertainty attention distillation. Conducting extensive evaluations on benchmark datasets, we demonstrate that our proposed uncertainty distillation is valuable to handle difficult labels of multi-label image classification.

Deep neural networks achieve remarkable performance in many computer vision tasks. Most state-of-the-art (SOTA) semantic segmentation and object detection approaches reuse neural network architectures designed for image classification as the backbone, commonly pre-trained on ImageNet. However, performance gains can be achieved by designing network architectures specifically for detection and segmentation, as shown by recent neural architecture search (NAS) research for detection and segmentation. One major challenge though is that ImageNet pre-training of the search space representation (a.k.a. super network) or the searched networks incurs huge computational cost. In this paper, we propose a Fast Network Adaptation (FNA++) method, which can adapt both the architecture and parameters of a seed network (e.g., an ImageNet pre-trained network) to become a network with different depths, widths, or kernel sizes via a parameter remapping technique, making it possible to use NAS for segmentation and detection tasks a lot more efficiently. In our experiments, we apply FNA++ on MobileNetV2 to obtain new networks for semantic segmentation, object detection, and human pose estimation that clearly outperform existing networks designed both manually and by NAS. We also implement FNA++ on ResNets and NAS networks, which demonstrates a great generalization ability. The total computation cost of FNA++ is significantly less than SOTA segmentation and detection NAS approaches: 1737× less than DPC, 6.8× less than Auto-DeepLab, and 8.0× less than DetNAS. A series of ablation studies are performed to demonstrate the effectiveness, and detailed analysis is provided for more insights into the working mechanism. Codes are available at https://github.com/JaminFong/FNA.

Face manipulation has shown remarkable advances with the flourish of Generative Adversarial Networks. However, due to the difficulties of controlling structures and textures, it is challenging to model poses and expressions simultaneously, especially for the extreme manipulation at high-resolution. In this article, we propose a novel framework that simplifies face manipulation into two correlated stages: a boundary prediction stage and a disentangled face synthesis stage. The first stage models poses and expressions jointly via boundary images. Specifically, a conditional encoder-decoder network is employed to predict the boundary image of the target face in a semi-supervised way. Pose and expression estimators are introduced to improve the prediction performance. In the second stage, the predicted boundary image and the input face image are encoded into the structure and the texture latent space by two encoder networks, respectively. A proxy network and a feature threshold loss are further imposed to disentangle the latent space. Furthermore, due to the lack of high-resolution face manipulation databases to verify the effectiveness of our method, we collect a new high-quality Multi-View Face (MVF-HQ) database. It contains 120,283 images at 6000 × 4000 resolution from 479 identities with diverse poses, expressions, and illuminations. MVF-HQ is much larger in scale and much higher in resolution than publicly available high-resolution face manipulation databases. We will release MVF-HQ soon to push forward the advance of face manipulation. Qualitative and quantitative experiments on four databases show that our method dramatically improves the synthesis quality.

Neural architecture search (NAS) has dramatically advanced the development of neural network design. We revisit the search space design in most previous NAS methods and find the number and widths of blocks are set manually. However, block counts and block widths determine the network scale (depth and width) and make a great influence on both the accuracy and the model cost (FLOPs/latency). In this paper, we propose to search block counts and block widths by designing a densely connected search space, i.e., DenseNAS. The new search space is represented as a dense super network, which is built upon our designed routing blocks. In the super network, routing blocks are densely connected and we search for the best path between them to derive the final architecture. We further propose a chained cost estimation algorithm to approximate the model cost during the search. Both the accuracy and model cost are optimized in DenseNAS. For experiments on the MobileNetV2-based search space, DenseNAS achieves 75.3% top-1 accuracy on ImageNet with only 361MB FLOPs and 17.9ms latency on a single TITAN-XP. The larger model searched by DenseNAS achieves 76.1% accuracy with only 479M FLOPs. DenseNAS further promotes the ImageNet classification accuracies of ResNet-18, -34 and -50-B by 1.5%, 0.5% and 0.3% with 200M, 600M and 680M FLOPs reduction respectively. The related code is available at https://github.com/JaminFong/DenseNAS.

Current state-of-the-art detectors typically exploit feature pyramid to detect objects at different scales. Among them, FPN is one of the representative works that build a feature pyramid by multi-scale features summation. However, the design defects behind prevent the multi-scale features from being fully exploited. In this paper, we begin by first analyzing the design defects of feature pyramid in FPN, and then introduce a new feature pyramid architecture named AugFPN to address these problems. Specifically, AugFPN consists of three components: Consistent Supervision, Residual Feature Augmentation, and Soft RoI Selection. AugFPN narrows the semantic gaps between features of different scales before feature fusion through Consistent Supervision. In feature fusion, ratio-invariant context information is extracted by Residual Feature Augmentation to reduce the information loss of feature map at the highest pyramid level. Finally, Soft RoI Selection is employed to learn a better RoI feature adaptively after feature fusion. By replacing FPN with AugFPN in Faster R-CNN, our models achieve 2.3 and 1.6 points higher Average Precision (AP) when using ResNet50 and MobileNet-v2 as backbone respectively. Furthermore, AugFPN improves RetinaNet by 1.6 points AP and FCOS by 0.9 points AP when using ResNet50 as backbone. Codes are available on https://github.com/Gus-Guo/AugFPN.

State-of-the-art pedestrian detectors have performed promisingly on non-occluded pedestrians, yet they are still confronted by heavy occlusions. Although many previous works have attempted to alleviate the pedestrian occlusion issue, most of them rest on still images. In this paper, we exploit the local temporal context of pedestrians in videos and propose a tube feature aggregation network (TFAN) aiming at enhancing pedestrian detectors against severe occlusions. Specifically, for an occluded pedestrian in the current frame, we iteratively search for its relevant counterparts along temporal axis to form a tube. Then, features from the tube are aggregated according to an adaptive weight to enhance the feature representations of the occluded pedestrian. Furthermore, we devise a temporally discriminative embedding module (TDEM) and a part-based relation module (PRM), respectively, which adapts our approach to better handle tube drifting and heavy occlusions. Extensive experiments are conducted on three datasets, Caltech, NightOwls and KAIST, showing that our proposed method is significantly effective for heavily occluded pedestrian detection. Moreover, we achieve the state-of-the-art performance on the Caltech and NightOwls datasets.

Despite the previous success of object analysis, detecting and segmenting a large number of object categories with a long-tailed data distribution remains a challenging problem and is less investigated. For a large-vocabulary classifier, the chance of obtaining noisy logits is much higher, which can easily lead to a wrong recognition. In this paper, we exploit prior knowledge of the relations among object categories to cluster fine-grained classes into coarser parent classes, and construct a classification tree that is responsible for parsing an object instance into a fine-grained category via its parent class. In the classification tree, as the number of parent class nodes are significantly less, their logits are less noisy and can be utilized to suppress the wrong/noisy logits existed in the fine-grained class nodes. As the way to construct the parent class is not unique, we further build multiple trees to form a classification forest where each tree contributes its vote to the fine-grained classification. To alleviate the imbalanced learning caused by the long-tail phenomena, we propose a simple yet effective resampling method, NMS Resampling, to re-balance the data distribution. Our method, termed as Forest R-CNN, can serve as a plug-and-play module being applied to most object recognition models for recognizing more than 1000 categories. Extensive experiments are performed on the large vocabulary dataset LVIS. Compared with the Mask R-CNN baseline, the Forest R-CNN significantly boosts the performance with 11.5% and 3.9% AP improvements on the rare categories and overall categories, respectively. Moreover, we achieve state-of-the-art results on the LVIS dataset. Code is available at https://github.com/JialianW/Forest_RCNN.

Detecting small-scale pedestrians is one of the most challenging problems in pedestrian detection. Due to the lack of visual details, the representations of small-scale pedestrians tend to be weak to be distinguished from background clutters. In this paper, we conduct an in-depth analysis of the small-scale pedestrian detection problem, which reveals that weak representations of small-scale pedestrians are the main cause for a classifier to miss them. To address this issue, we propose a novel Self-Mimic Learning (SML) method to improve the detection performance on small-scale pedestrians. We enhance the representations of small-scale pedestrians by mimicking the rich representations from large-scale pedestrians. Specifically, we design a mimic loss to force the feature representations of small-scale pedestrians to approach those of large-scale pedestrians. The proposed SML is a general component that can be readily incorporated into both one-stage and two-stage detectors, with no additional network layers and incurring no extra computational cost during inference. Extensive experiments on both the CityPersons and Caltech datasets show that the detector trained with the mimic loss is significantly effective for small-scale pedestrian detection and achieves state-of-the-art results on CityPersons and Caltech, respectively.

Synthetic visual data refers to the data automatically rendered by the mature computer graphic algorithms. With the rapid development of these techniques, we can now collect photo-realistic synthetic images with accurate pixel-level annotations without much effort. However, due to the domain gaps between synthetic data and real data, in terms of not only visual appearance but also label distribution, directly applying models trained on synthetic images to real ones can hardly yield satisfactory performance. Since the collection of accurate labels for real images is very laborious and time-consuming, developing algorithms which can learn from synthetic images is of great significance. In this paper, we propose a novel framework, namely Active Pseudo-Labeling (APL), to reduce the domain gaps between synthetic images and real images. In APL framework, we first predict pseudo-labels for the unlabeled real images in the target domain by actively adapting the style of the real images to source domain. Specifically, the style of real images is adjusted via a novel task guided generative model, and then pseudo-labels are predicted for these actively adapted images. Lastly, we fine-tune the source-trained model in the pseudo-labeled target domain, which helps to fit the distribution of the real data. Experiments on both semantic segmentation and object detection tasks with several challenging benchmark data sets demonstrate the priority of our proposed method compared to the existing state-of-the-art approaches.

Deep convolutional neural networks have achieved great success on various image recognition tasks. However, it is nontrivial to transfer the existing networks to video due to the fact that most of them are developed for static image. Frame-byframe processing is suboptimal because temporal information that is vital for video understanding is totally abandoned. Furthermore, frame-by-frame processing is slow and inefficient, which can hinder the practical usage. In this paper, we propose LWDN (Locally-Weighted Deformable Neighbors) for video object detection without utilizing time-consuming optical flow extraction networks. LWDN can latently align the high-level features between keyframes and keyframes or nonkeyframes. Inspired by (Zhu et al. 2017a) and (Hetang et al. 2017) who propose to aggregate features between keyframes and keyframes, we adopt brain-inspired memory mechanism to propagate and update the memory feature from keyframes to keyframes. We call this process Memory-Guided Propagation. With such a memory mechanism, the discriminative ability of features in keyframes and non-keyframes are both enhanced, which helps to improve the detection accuracy. Extensive experiments on VID dataset demonstrate that our method achieves superior performance in a speed and accuracy trade-off, i.e., 76.3% on the challenging VID dataset while maintaining 20fps in speed on Titan X GPU.

Recent years have witnessed the great success of deep learning models in semantic segmentation. Nevertheless, these models may not generalize well to unseen image domains due to the phenomenon of domain shift. Since pixel-level annotations are laborious to collect, developing algorithms which can adapt labeled data from source domain to target domain is of great significance. To this end, we propose self-ensembling attention networks to reduce the domain gap between different datasets. To the best of our knowledge, the proposed method is the first attempt to introduce selfensembling model to domain adaptation for semantic segmentation, which provides a different view on how to learn domain-invariant features. Besides, since different regions in the image usually correspond to different levels of domain gap, we introduce the attention mechanism into the proposed framework to generate attention-aware features, which are further utilized to guide the calculation of consistency loss in the target domain. Experiments on two benchmark datasets demonstrate that the proposed framework can yield competitive performance compared with the state of the art methods.

Neural Architecture Search (NAS) is an important yet challenging task in network design due to its high computational consumption. To address this issue, we propose the Reinforced Evolutionary Neural Architecture Search (RENAS), which is an evolutionary method with reinforced mutation for NAS. Our method integrates reinforced mutation into an evolution algorithm for neural architecture exploration, in which a mutation controller is introduced to learn the effects of slight modifications and make mutation actions. The reinforced mutation controller guides the model population to evolve efficiently. Furthermore, as child models can inherit parameters from their parents during evolution, our method requires very limited computational resources. In experiments, we conduct the proposed search method on CIFAR-10 and obtain a powerful network architecture, RENASNet. This architecture achieves a competitive result on CIFAR-10. The explored network architecture is transferable to ImageNet and achieves a new state-of-the-art accuracy, i.e., 75.7% top-1 accuracy with 5.36M parameters on mobile ImageNet. We further test its performance on semantic segmentation with DeepLabv3 on the PASCAL VOC. RENASNet outperforms MobileNet-v1, MobileNet-v2 and NASNet. It achieves 75.83% mIOU without being pretrained on COCO.

Transferring image-based object detectors to the domain of videos remains a challenging problem. Previous efforts mostly exploit optical flow to propagate features across frames, aiming to achieve a good trade-off between accuracy and efficiency. However, introducing an extra model to estimate optical flow can significantly increase the overall model size. The gap between optical flow and high-level features can also hinder it from establishing spatial correspondence accurately. Instead of relying on optical flow, this paper proposes a novel module called Progressive Sparse Local Attention (PSLA), which establishes the spatial correspondence between features across frames in a local region with progressively sparser stride and uses the correspondence to propagate features. Based on PSLA, Recursive Feature Updating (RFU) and Dense Feature Transforming (DenseFT) are proposed to model temporal appearance and enrich feature representation respectively in a novel video object detection framework. Experiments on ImageNet VID show that our method achieves the best accuracy compared to existing methods with smaller model size and acceptable runtime speed.

Person Re-identification (ReID) is a challenging retrieval task that requires matching a person's image across non-overlapping camera views. The quality of fulfilling this task is largely determined on the robustness of the features that are used to describe the person. In this paper, we show the advantage of jointly utilizing multi-scale abstract information to learn powerful features over full body and parts. A scale normalization module is proposed to balance different scales through residual-based integration. To exploit the information hidden in non-rigid body parts, we propose an anchor-based method to capture the local contents by stacking convolutions of kernels with various aspect ratios, which focus on different spatial distributions. Finally, a well-defined framework is constructed for simultaneously learning the representations of both full body and parts. Extensive experiments conducted on current challenging large-scale person ReID datasets, including Market1501, CUHK03 and DukeMTMC, demonstrate that our proposed method achieves the state-of-the-art results.

In hyperspectral remote sensing data mining, it is important to take into account of both spectral and spatial information, such as the spectral signature, texture feature, and morphological property, to improve the performances, e.g., the image classification accuracy. In a feature representation point of view, a nature approach to handle this situation is to concatenate the spectral and spatial features into a single but high dimensional vector and then apply a certain dimension reduction technique directly on that concatenated vector before feed it into the subsequent classifier. However, multiple features from various domains definitely have different physical meanings and statistical properties, and thus such concatenation has not efficiently explore the complementary properties among different features, which should benefit for boost the feature discriminability. Furthermore, it is also difficult to interpret the transformed results of the concatenated vector. Consequently, finding a physically meaningful consensus low dimensional feature representation of original multiple features is still a challenging task. In order to address these issues, we propose a novel feature learning framework, i.e., the simultaneous spectral-spatial feature selection and extraction algorithm, for hyperspectral images spectral-spatial feature representation and classification. Specifically, the proposed method learns a latent low dimensional subspace by projecting the spectral-spatial feature into a common feature space, where the complementary information has been effectively exploited, and simultaneously, only the most significant original features have been transformed. Encouraging experimental results on three public available hyperspectral remote sensing datasets confirm that our proposed method is effective and efficient.

Low-rank matrix approximation has been widely used for data subspace clustering and feature representation in many computer vision and pattern recognition applications. However, in order to enhance the discriminability, most of the matrix approximation based feature extraction algorithms usually generate the cluster labels by certain clustering algorithm (e.g., the kmeans) and then perform the matrix approximation guided by such label information. In addition, the noises and outliers in the dataset with large reconstruction errors will easily dominate the objective function by the conventional ℓ2-norm based squared residue minimization. In this paper, we propose a novel clustering and feature extraction algorithm based on an unified low-rank matrix factorization framework, which suggests that the observed data matrix can be approximated by the production of projection matrix and low dimensional representation, among which the low-dimensional representation can be approximated by the cluster indicator and latent feature matrix simultaneously. Furthermore, we have proposed using the ℓ2,1-norm and integrating the manifold regularization to further promote the proposed model. A novel Augmented Lagrangian Method (ALM) based procedure is designed to effectively and efficiently seek the optimal solution of the problem. The experimental results in both clustering and feature extraction perspectives demonstrate the superior performance of the proposed method.

Data clustering is the task to group the data samples into certain clusters based on the relationships of samples and structures hidden in data, and it is a fundamental and important topic in data mining and machine learning areas. In the literature, the spectral clustering is one of the most popular approaches and has many variants in recent years. However, the performance of spectral clustering is determined by the affinity matrix, which is always computed by a predefined model (e.g., Gaussian kernel function) with carefully tuned parameters combination, and may far from optimal in practice. In this paper, we propose to consider the observed data clustering as a robust matrix factorization point of view, and learn an affinity matrix simultaneously to regularize the proposed matrix factorization. The solution of the proposed adaptive manifold regularized matrix factorization (AMRMF) is reached by a novel Augmented Lagrangian Multiplier (ALM) based algorithm. The experimental results on standard clustering datasets demonstrate the superior performance over the exist alternatives.

In recent years, various data clustering algorithms have been proposed in the data mining and engineering communities. However, there are still drawbacks in traditional clustering methods which are worth to be further investigated, such as clustering for the high dimensional data, learning an ideal affinity matrix which optimally reveals the global data structure, discovering the intrinsic geometrical and discriminative properties of the data space, and reducing the noises influence brings by the complex data input. In this paper, we propose a novel clustering algorithm called robust dual clustering with adaptive manifold regularization (RDC), which simultaneously performs dual matrix factorization tasks with the target of an identical cluster indicator in both of the original and projected feature spaces, respectively. Among which, the$l_{2,1}$-norm is used instead of the conventional$l_{2}$-norm to measure the loss, which helps to improve the model robustness by relieving the influences by the noises and outliers. In order to better consider the intrinsic geometrical and discriminative data structure, we incorporate the manifold regularization term on the cluster indicator by using a particularly learned affinity matrix which is more suitable for the clustering task. Moreover, a novel augmented lagrangian method (ALM) based procedure is designed to effectively and efficiently seek the optimal solution of the proposed RDC optimization. Numerous experiments on the representative data sets demonstrate the superior performance of the proposed method compares to the existing clustering algorithms.

Pattern recognition tasks such as the data classification and clustering usually can be represented by the perspective of multiple views or feature spaces. Obviously, the accuracy of the classification and clustering should be greatly improved if we carefully consider the discriminabilities from multiple views and explore the complementary information among them. However, multiple features also bring new challenges to handle them. In the literature, many existed multiview feature learning methods dealt with different views equally, thus they couldn't optimally utilize the complementary property of them. On the other hand, the matrix factorization based clustering algorithms usually adopt the conventional ℓ2-norm based squared residue minimization to measure the loss, which is easily influenced by the outliers and noises from the multiple sources of input. In this paper, we propose a novel multiview data clustering algorithm based on the matrix factorization to relieve the above issues. The basic idea of the proposed Robust and Regularized Matrix Approximation (RRMA) is that the observed data matrix could be low-rank approximated by a cluster centroid matrix and a cluster indicator matrix, respectively, and the major contributions of our work lie in the introduction of the robust ℓ2,1-norm and ensemble manifold regularization to regularize the matrix factorization and make the model more discriminative for multiview data clustering. We properly adjust the importance of different views by assigning a set of trainable weights on the views. Moreover, we propose an efficient solution featured with impactful updating rules to seek the local optimal parameters. Encouraging experimental results on numerous public multiview datasets demonstrate the superiority of our model compared to some state-of-the-art methods.

In data mining and pattern recognition area, the learned objects are often represented by the multiple features from various of views. How to learn an efficient and effective feature embedding for the subsequent learning tasks? In this paper, we address this issue by providing a novel multi-task multiview feature embedding (MMFE) framework. The MMFE algorithm is based on the idea of low-rank approximation, which suggests that the observed multiview feature matrix is approximately represented by the low-dimensional feature embedding multiplied by a projection matrix. In order to fully consider the particular role of each view to the multiview feature embedding, we simultaneously suggest the multitask learning scheme and ensemble manifold regularization into the MMFE algorithm to seek the optimal projection. Since the objection function of MMFE is multi-variable and non-convex, we further provide an iterative optimization procedure to find the available solution. Two real world experiments show that the proposed method outperforms single-task-based as well as state-of-the-art multiview feature embedding methods for the classification problem.

Spectral-spatial feature combination for hyperspectral image analysis has become an important research topic in hyperspectral remote sensing applications. A simple and straightforward way to integrate spectral-spatial features is to concatenate heterogeneous features into a long vector. Then, the dimensionality reduction techniques, i.e., feature selection, are applied before subsequent utilizations. However, such representation can introduce redundancy and noise. Moreover, traditional single-feature selection methods treat different features equally and ignore their complementary properties. As a result, the performance of subsequent tasks, i.e., classification, would drop down. In this paper, we propose a novel approach to integrate the spectral-spatial features based on the concatenating strategy, termed discriminative sparse multimodal learning for feature selection (DSML-FS). In the proposed method, joint structured sparsity regularizations are used to exploit the intrinsic data structure and relationships among different features. Discriminative least squares regression is applied to enlarge the distance between classes. Therefore, the weight matrix incorporating the information of feature wise and individual properties is automatically learned for spectral-spatial feature selection. We develop an alternative iterative algorithm to solve the nonlinear optimization problem in DSML-FS with global convergence. We systematically evaluate the proposed algorithm on three available hyperspectral data sets, and the encouraging experimental results demonstrate the effectiveness of DSML-FS.

In hyperspectral image (HSI) classification, feature extraction is one important step. Traditional methods, e.g., principal component analysis (PCA) and locality preserving projection, usually neglect the information of within-class similarity and between-class dissimilarity, which is helpful to the improvement of classification. On the other hand, most of these methods, e.g., PCA and linear discriminative analysis, consider that the HSI data lie on a low-dimensional manifold or each class is on a submanifold. However, some class data of HSI may lie on a multimanifold. To avoid these problems, we propose a method for feature extraction in HSIs, assuming that a local region resides on a submainfold. In our method, we deal with the data region by region by taking into account the different discriminative locality information. Then, under the metric learning framework, a robust distance metric is learned. It aims to learn a subspace in which the samples in the same class are as near as possible while the samples in different classes are as far as possible. Encouraging experimental results on two available hyperspectral data sets indicate that our proposed algorithm outperforms many existing feature extract methods for HSI classification.