Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Chi–Keung TangSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works232 from public data
- RMPE: Regional Multi-person Pose Estimation1,884
This paper proposes a novel regional multi-person pose estimation (RMPE) framework to facilitate pose estimation in the presence of inaccurate human bounding boxes and can achieve 76:7 mAP on the MPII (multi person) dataset.
- Segment Anything in High Quality714
A learnable High-Quality Output Token is injected into SAM's mask decoder and is responsible for predicting the high-quality mask, which reuses and preserves the pre-trained model weights of SAM, while only introducing minimal additional parameters and computation.
- 669
- Few-Shot Object Detection With Attention-RPN and Multi-Relation Detector648
A novel few-shot object detection network that aims at detecting objects of unseen categories with only a few annotated examples, which exploits the similarity between the few shot support set and query set to detect novel objects while suppressing false detection in the background.
- 541
- 497
- Deep Saliency with Encoded Low Level Distance Map and High Level Features487
It is demonstrated that hand-crafted features can provide complementary information to enhance performance of saliency detection that utilizes only high level features.
- 410
- 370
- 354
- FSS-1000: A 1000-Class Dataset for Few-Shot Segmentation339
This paper builds a few-shot segmentation dataset, FSS-1000, which consists of 1000 object classes with pixelwise annotation of ground-truth segmentation and finds that training the model from scratch using F SS-1000 achieves comparable and even better results than training with weights pre-trained by ImageNet.
- 334
- Deep High Dynamic Range Imaging with Large Foreground Motions311
This paper proposes the first non-flow-based deep framework for high dynamic range (HDR) imaging of dynamic scenes with large-scale foreground motions, and produces excellent results where color artifacts and geometric distortions are significantly reduced compared to existing state-of-the-art methods.
- Accurate Single Stage Detector Using Recurrent Rolling Convolution297
A novel single stage end-to-end trainable object detection network is proposed by introducing Recurrent Rolling Convolution (RRC) architecture over multi-scale feature maps to construct object classifiers and bounding box regressors which are deep in context.
- 282
- CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation via Global and Local Refinement276
This paper presents quantitative and qualitative studies on different datasets to show that CascadePSP can reveal pixel-accurate segmentation boundaries using the novel refinement module without any finetuning.
- Semantic colorization with internet images271
A colorization system that leverages the rich image content on the internet and the user needs only to provide a semantic text label and segmentation cues for major foreground objects in the scene to achieve the desired result.
- Modular Interactive Video Object Segmentation: Interaction-to-Mask, Propagation and Difference-Aware Fusion253
The MiVOS framework is presented, which decouples interaction-to-mask and mask propagation, allowing for higher generalizability and better performance, and a large-scale synthetic VOS dataset with pixel-accurate segmentation of 4.8M frames is contributed to facilitate future research.
- Memory-Attended Recurrent Network for Video Captioning240
The Memory-Attended Recurrent Network (MARN) for video captioning is proposed, in which a memory structure is designed to explore the full-spectrum correspondence between a word and its various similar visual contexts across videos in training data.
- Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers239
The explicit modeling of occlusion relationship with bilayer structure naturally decouples the boundaries of both the occluding and occluded instances, and considers the interaction between them during mask regression, to validate the efficacy of bilayer decoupling on both one-stage and two-stage object detectors with different backbones and network layer choices.
- Partial Sum Minimization of Singular Values in Robust PCA: Algorithm and Applications239
Instead of minimizing the nuclear norm, this paper proposes to minimize the partial sum of singular values, which implicitly encourages the target rank constraint, and shows that its results outperform those obtained by the conventional nuclear norm rank minimization method.
- Cross-Domain Adaptation for Animal Pose Estimation227
This paper proposed a novel cross-domain adaptation method to transform the animal pose knowledge from labeled animal classes to unlabeled animal classes, and uses the modest animal pose dataset to adapt learned knowledge to multiple animals species.
- 226
- Attribute-Guided Face Generation Using Conditional CycleGAN224
This work condition the CycleGAN and proposes conditional CycleGAN, which is designed to handle unpaired training data because the training low/high-res and high-res attribute images may not necessarily align with each other, and to allow easy control of the appearance of the generated face via the input attributes.
- Learning Dual Convolutional Neural Networks for Low-Level Vision221
The proposed DualCNN consists of two parallel branches, which respectively recovers the structures and details in an end-to-end manner, and can generate the target signals according to the formation model for each particular application.
- Dense Hybrid Recurrent Multi-view Stereo Net with Dynamic Consistency Checking216
An efficient and effective dense hybrid recurrent multi-view stereo net with dynamic consistency checking, namely D^{2}$HC-RMVSNet, for accurate dense point cloud reconstruction and dynamically aggregate geometric consistency matching error among all the views is proposed.
- 202
- 202
- 200
- Physics-Based Generative Adversarial Models for Image Restoration and Beyond191
An algorithm to directly solve numerous image restoration problems (e.g., image deblurring, image dehazing, and image deraining) by generative models with adversarial learning within the GAN framework.
- Learning Video Object Segmentation From Unlabeled Videos184
A unified unsupervised/weakly supervised learning framework, called MuG, that comprehensively captures intrinsic properties of VOS at multiple granularities is introduced that can help advance understanding of visual patterns in VOS and significantly reduce annotation burden.
- Natural Image Matting Using Deep Convolutional Neural Networks184
A deep Convolutional Neural Networks method for natural image matting that can “recognize” different local image structures, and then combine results of local (closed form matting), and nonlocal (KNN matting) matting effectively to achieve higher quality alpha mattes than both of its inputs.
- Mask Transfiner for High-Quality Instance Segmentation173
This paper presents Mask Transfiner, a transformer-based approach to instance segmentation that decomposes and represents the image regions as a quadtree, allowing it to predict highly accurate instance masks, at a low computational cost.
- NeRF-SR: High Quality Neural Radiance Fields using Supersampling167
The method is built upon Neural Radiance Fields (NeRF) that predicts per-point density and color with a multi-layer perceptron and can be exploited by a super-sampling strategy that shoots multiple rays at each image pixel, which further enforces multi-view constraint at a sub-pixel level.
- Adversarial Attacks Beyond the Image Space165
Though image-space adversaries can be interpreted as per-pixel albedo change, it is verified that they cannot be well explained along these physically meaningful dimensions, which often have a non-local effect.
- 159
- Image Generation from Sketch Constraint Using Contextual GAN153
This paper addresses image generation guided by hand sketch using a novel joint image completion approach, where the sketch provides the image context for completing, or generating the output image, using a generated adversarial network.
- 151
- Pairwise Body-Part Attention for Recognizing Human-Object Interactions148
A new pairwise body-part attention model which can learn to focus on crucial parts, and their correlations for HOI recognition is proposed, and a novel attention based feature selection method and a feature representation scheme that can capture pairwise correlations between body parts are introduced.
- 145
- 145
- Pyramid Multi-view Stereo Net with Self-adaptive View Aggregation141
This paper proposes an effective and efficient pyramid multi-view stereo (MVS) net with self-adaptive view aggregation for accurate and complete dense point cloud reconstruction and establishes a new state-of-the-art on the DTU dataset with significant improvements in the completeness and overall quality.
- Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation132
- LADN: Local Adversarial Disentangling Network for Facial Makeup and De-Makeup131
A local adversarial disentangling network for facial makeup and de-makeup that can distinguish whether the generated local image details are consistent with the corresponding regions in the given reference image in cross-image style transfer in an unsupervised setting is proposed.
- 131
- A Unified Approach of Multi-scale Deep and Hand-Crafted Features for Defocus Estimation130
This paper systematically analyzes the effectiveness of different features, and shows how each feature can compensate for the weaknesses of other features when they are concatenated.
- Weakly and Semi Supervised Human Body Part Parsing via Pose-Guided Knowledge Transfer128
A novel method to generate synthetic human part segmentation data using easily-obtained human keypoint annotations to exploit the anatomical similarity among human to transfer the parsing results of a person to another person with similar pose.
- 122
- 122
- 120
- Segment Anything Meets Point Tracking119
This paper introduces SAM-PT, an innovative method for point-centric video segmentation, leveraging the capabilities of SAM alongside long-term point tracking, and uniquely exploits point propagation to utilize local structure information independent of object semantics.
- 118
- Look, Listen and Learn — A Multimodal LSTM for Speaker Identification115
A novel multimodal Long Short-Term Memory (LSTM) architecture which seamlessly unifies both visual and auditory modalities from the beginning of each sequence input is described which outperforms the state-of-the-art systems in speaker identification with lower false alarm rate and higher recognition accuracy.
- Semantic Image Matting111
This work considers and learns 20 classes of matting patterns, and proposes to extend the conventional trimap to semantic trimap, and adds a multi-class discriminator to regularize the alpha prediction at semantic level, and content-sensitive weights to balance different regularization losses.
- 110
- Reflective Decoding Network for Image Captioning109
It is shown that vocabulary coherence between words and syntactic paradigm of sentences are also important to generate high-quality image captioning, and the proposed Reflective Decoding Network (RDN) enhances both the long-sequence dependency and position perception of words in a caption decoder.
- Group Collaborative Learning for Co-Salient Object Detection105
This paper’s new technical contributions on a number of important downstream computer vision applications including content aware co-segmentation, co-localization based automatic thumbnails, etc are demonstrated.
- 105
- 100
- 97
- 92
- NeRF-RPN: A general framework for object detection in NeRFs89
This paper presents the first significant object detection framework, NeRF-RPN, which directly operates on NeRF, and demonstrates it is possible to regress the 3D bounding boxes of objects in NeRF directly without rendering the NeRF at any viewpoint.
- 88
- 88
- Weakly- and Self-Supervised Learning for Content-Aware Deep Image Retargeting83
This paper proposes a weakly- and self-supervised deep convolutional neural network (WSSDCNN) for contentaware image retargeting, which takes a source image and a target aspect ratio, and then directly outpues a retargeted image.
- Conditional CycleGAN for Attribute Guided Face Image Generation80
The method to incorporate Z into the network is presented, such that the hallucinated high resolution face image $Y'$ not only satisfies the low resolution constrain inherent in $X$, but also the attribute condition prescribed by $Z$.
- Deep Video Matting via Spatio-Temporal Alignment and Aggregation78
Quantitative and qualitative experimental results show that the ST-FAM framework significantly outperforms conventional video matting and deep image matting methods applied to video in presence of multi-frame temporal information.
- 78
- Fast Video Object Segmentation With Temporal Aggregation Network and Dynamic Template Matching75
“tracking-by-detection” is introduced into VOS which can coherently integrates segmentation into tracking, by proposing a new temporal aggregation network and a novel dynamic time-evolving template matching mechanism to achieve significantly improved performance.
- Pointwise Rotation-Invariant Network with Adaptive Sampling and 3D Spherical Voxel Convolution75
A brand new point-set learning framework PRIN is proposed, namely, Pointwise Rotation-Invariant Network, focusing on rotation-invariant feature extraction in point clouds analysis, which shows that, on the dataset with randomly rotated point clouds, PRIN demonstrates better performance than state-of-the-art methods without any data augmentation.
- 73
- Fast randomized Singular Value Thresholding for Nuclear Norm Minimization71
This paper proposes an accurate and fast approximation method for SVT, called fast randomized SVT (FRSVT), where the key idea is to extract an approximate basis for the range of a matrix from its compressed matrix.
- 70
- Cascade-DETR: Delving into High-Quality Universal Object Detection69
Cascade-DETR is introduced for high-quality universal object detection, and the Cascade Attention layer is proposed, which explicitly integrates object-centric information into the detection decoder by limiting the attention to the previous box prediction.
- 68
- HAA500: Human-Centric Atomic Action Dataset with Curated Videos67
The extensive experiments including cross-data validation using datasets collected in the wild demonstrate the clear benefits of human-centric and atomic characteristics of HAA500, which enable training even a baseline deep learning model to improve prediction by attending to atomic human poses.
- Interactiveness Field in Human-Object Interactions64
This work introduces a previously overlooked interactiveness bimodal prior, and proposes new energy constraints based on the cardinality and difference in the inherent “interactiveness field” underlying interactive versus non-interactive pairs that can detect more precise pairs and significantly boost HOI detection performance.
- 64
- Look Back and Forth: Video Super-Resolution with Explicit Temporal Difference Modeling62
This paper proposes to compute the temporal difference between frames and divide those pixels into two subsets according to the level of difference and demonstrates the effectiveness of the proposed method and its favorable performance against state-of-the-art methods.
- 62
- Dual Convolutional Neural Networks for Low-Level Vision59
Experimental results show that the DualCNN can be effectively applied to numerous low-level vision tasks with favorable performance against the state-of-the-art methods that have been specially designed for each individual task.
- 59
- GSNet: Joint Vehicle Pose and Shape Reconstruction with Geometrical and Scene-Aware Supervision56
A novel end-to-end framework named GSNet (Geometric and Scene-aware Network), which jointly estimates 6DoF poses and reconstructs detailed 3D car shapes from single urban street view, and inspires a new multi-objective loss function to regularize network training.
- 55
- 54
- Instance Neural Radiance Field51
Instance-NeRF surpasses previous NeRF segmentation works and competitive 2D segmentation methods in segmentation performance on unseen views and is also one of the first to achieve such results in pure inference.
- 48
- 48
- Occlusion-Aware Video Object Inpainting47
Experimental results clearly demonstrate the efficacy of the occlusion-aware video object inpainting, which recovers both the complete shape and appearance for occluded objects in videos given their visible mask segmentation.
- 47
- GCoNet+: A Stronger Group Collaborative Co-Salient Object Detector46
A novel end-to-end group collaborative learning network, termed GCoNet+, which can effectively and efficiently identify co-salient objects in natural scenes and outperforms the existing 12 cutting-edge models.
- 46
- Registration Based Non‐uniform Motion Deblurring46
An algorithm which uses image registration to estimate a non‐uniform motion blur point spread function (PSF) caused by camera shake is proposed and the effectiveness and limitations of the algorithm for non‐ uniform deblurring are discussed.
- 45
- 44
- 42
- 41
- Video Matting Using Multi-frame Nonlocal Matting Laplacian41
The approach extends the conventional image matting approach to video by using multi-frame nonlocal matting Laplacian, and adopts the nonlocal mean regularization to enhance temporal coherence of the estimated alpha mattes and to correct alpha matte errors at low contrast regions.
- 40
- Refining Geometry from Depth Sensors using IR Shading Images39
A near light source IR shading model is defined that describes the captured intensity as a function of surface normals, albedo, lighting direction, and distance between light source and surface points, to resolve the ambiguity in the model between the normals and distances.
- Occlusion-Aware Instance Segmentation Via BiLayer Network Architectures37
Large and consistent improvements using one/two-stage and query-based object detectors with various backbones and network layer choices validate the generalization ability of bilayer decoupling, as shown by extensive experiments on image instance segmentation benchmarks.
- 36
- Consistent Matting for Light Field Images36
A new image matting algorithm is presented to extract consistent alpha mattes across sub-images of a light field image by utilizing the epipolar plane image (EPI) to construct comprehensive foreground and background sample sets across the sub- images without missing a true sample.
- 36
- 35
- Mask-Free Video Instance Segmentation32
This work proposes MaskFreeVIS, achieving highly competitive VIS performance, while only using bounding box annotations for the object state, by introducing the Temporal KNN-patch Loss (TK-Loss), providing strong mask supervision without any labels.
- 32
- Richardson–Lucy deblurring for scenes under a projective motion path32
A new Projective Motion Blur Model that treats the blurred image as an integration of a clear scene under a sequence of projective transformations that describe the camera’s path and represents the degraded image as an integration of the clear scene under a sequence of projective motions.
- 30
- 29
- 28
- 28
- 28
- 28
- 27
- 26
- 26
- 24
- 24
- 24
- 24
- PRIN: Pointwise Rotation-Invariant Network.23
A new point sets learning framework named Pointwise Rotation-Invariant Network (PRIN), focusing on the rotation problem in point clouds, which shows similar performance on par or better than state-of-the-art methods on part segmentation without data augmentation.
- 23
- 21
- 20
- 18
- 18
- 18
- 18
- 17
- 16
- 15
- 15
- 15
- 15
- 15
- 14
- 14
- 14
- 14
- 13
- 12
- 12
- 12
- 11
- 11
- 11
- 11
- 9
- 9
- 9
- 9
- 9
- 8
- 8
- 8
- 7
- 7
- 7
- 6
- 6
- 6
- 6
- 6
- 6
- 5
- 5
- 5
- 5
- 5
- 5
- 5
- 4
- 4
- 4
- 4
- 4
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-13. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.