Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks116 from public data
- Towards Efficient and Scalable Sharpness-Aware Minimization183
A novel algorithm LookSAM is proposed - that only periodically calculates the inner gradient ascent, to significantly reduce the additional training cost of SAM and enjoys comparable computational complexity with first-order optimizers such as SGD or Adam.
- E-NeRV: Expedite Neural Video Representation with Disentangled Spatial-Temporal Context145
This paper proposes E-NeRV, which dramatically expedites NeRV by decomposing the image-wise implicit neural representation into separate spatial and temporal context, and greatly reduces the redundant model parameters, while retaining the representation ability.
- Go Wider Instead of Deeper112
This paper designs WideNet, a parameter-efficient framework that scales along model width by replacing feed-forward network with mixture-of-experts (MoE), and proposes to use individual layernorms to transform various semantic representations in a more parameter- efficient way.
- Designing One Unified Framework for High-Fidelity Face Reenactment and Swapping74
This paper proposes an effective end-to-end unified framework to achieve both face reenactment and swapping, and sufficiently transfers identity and attribute based on learned disentangled representations to generate high-fidelity faces.
- Adaptive Assignment for Geometry Aware Local Feature Matching68
AdaMatcher is in-troduce, which first accomplishes the feature correlation and co-visible area estimation through an elaborate feature interaction module, then performs adaptive assignment on patch-level matching while es-timating the scales between images, and finally refines the co- visible matches through scale alignment and sub-pixel regression module.
- UltraVideo: High-Quality UHD Video Dataset with Comprehensive Captions59
This paper proposes a high-quality open-sourced UHD-4K text-to-video dataset named UltraVideo, and expands Wan to UltraWan-1K/-4K, which can natively generate high-quality 1K/4K videos with more consistent text controllability, demonstrating the effectiveness of the data curation.
- Revisiting the Spatial and Temporal Modeling for Few-shot Action Recognition58
This paper proposes SloshNet, a new framework that revisits the spatial and temporal modeling for few-shot action recognition in a finer manner, and introduces a long-term temporal modeling module to model the global temporal relations based on the extracted spatial appearance features.
- 49
- Medical Phrase Grounding with Region-Phrase Context Contrastive Alignment47
Experimental results on three MPG datasets demonstrate that the proposed MedRPG outperforms state-of-the-art visual grounding approaches by a large margin, and the proposed TaCo strategy is effective in enhancing finding localization ability and reducing spurious region-phrase correlations.
- Global Spectral Filter Memory Network for Video Object Segmentation44
This paper proposes Global Spectral Filter Memory network (GSFM), which improves intra-frame interaction through learning long-term spatial dependencies in the spectral domain and proposes Low (High) Frequency Module, which is proposed to fit this circumstance.
- Guide Local Feature Matching by Overlap Estimation41
A novel Overlap Estimation method conditioned on image pairs with TRansformer, named OETR, to constrain local feature matching in the commonly visible region, and can boost state of the art local feature Matching performance substantially.
- 38
- RIFormer: Keep Your Vision Backbone Effective But Removing Token Mixer33
This paper develops an RepIdentityFormer base on the re-parameterizing idea, and explores the improved learning paradigm to break the limitation of simple token mixer free backbone, and summarizes the empirical practice into 5 guidelines.
- TIMotion: Temporal and Interactive Framework for Efficient Human-Human Motion Generation31
This paper introduces TIMotion (Temporal and Interactive Modeling), an efficient and effective framework for human-human motion generation and proposes Causal Interactive Injection to model two separate sequences as a causal sequence leveraging the temporal and causal properties.
- Thought-Augmented Planning for LLM-Powered Interactive Recommender Agent27
TAIRA is designed as an LLM-powered multi-agent system featuring a manager agent that orchestrates recommendation tasks by decomposing user needs and planning subtasks, with its planning capacity strengthened through Thought Pattern Distillation (TPD), a thought-augmentation method that extracts high-level thoughts from the agent's and human experts' experiences.
- Computer vision-based gesture recognition for an augmented reality interface27
This paper presents a computer vision-based gesture interface that is part of an Augmented Reality system that can recognise a 3D pointing gesture, a click gesture, and five static gestures.
- 24
- 23
- 23
- 20
- 19
- M2-CLIP: A Multimodal, Multi-task Adapting Framework for Video Action Recognition17
This work designs a multi-task decoder with a rich set of supervisory signals, including the original contrastive learning head, a cross-modal classification head, a cross-modal masked language modeling head, and a visual classification head that adeptly satisfies the need for strong supervised performance within a multimodal framework.
- Multimodal-driven Talking Face Generation via a Unified Diffusion-based Generator17
This work derives a novel paradigm free of unstable seesaw-style optimization, resulting in simple, stable, and effective training and inference schemes, and adopts the Diffusion Models.
- Multiple Object Tracking of Drone Videos by a Temporal-Association Network with Separated-Tasks Structure16
By combing the temporal-association structure and separating different function heads, the model’s performance in object detection and tracking tasks is improved and has better tracking performance than the models such as SORT and HDHNet on the UAV video dataset.
- 15
- 15
- A Generalist FaceX via Learning Unified Facial Representation14
This work presents FaceX framework, a novel facial generalist model capable of handling diverse facial tasks simultaneously, and introduces Facial Omni-Representation Decomposing (FORD) for seamless manipulation of various facial components, microscopically decomposing the core aspects of most facial editing tasks.
- Flat mirror tilt and piston measurement based on structured light reflection14
A simple method that can measure the flat mirror tilt for about 1″ accuracy and piston for less than 1um accuracy and under the guidance of the method, a mirror can be adjusted to a specified pose.
- 13
- AdaCM: Adaptive ColorMLP for Real-Time Universal Photo-realistic Style Transfer12
This work proposes the Adaptive ColorMLP (AdaCM), an effective and efficient framework for universal photo-realistic style transfer that is ultrafast and can process a 4K resolution image in 6ms on one V100 GPU.
- High-precision dynamic three-dimensional shape measurement of specular surfaces based on deep learning11
It is demonstrated for, what is believed to be, the first time that deep learning techniques can be combined with dynamic-PMD and can be used to reconstruct high-precision 3D shapes of specular surfaces from single-frame distorted orthogonal fringe patterns, enabling high-quality dynamic measurement of SPECular objects.
- 11
- 11
- 11
- Gaussian-LIC2: LiDAR-Inertial-Camera Gaussian Splatting SLAM10
This paper presents the first photo-realistic LiDAR-Inertial-Camera Gaussian Splatting SLAM system that simultaneously addresses visual quality, geometric accuracy, and real-time performance, and explores how the incrementally reconstructed Gaussian map can improve the robustness of odometry.
- Semi-supervised medical image segmentation with diverse views via cross-pseudo supervision9
Compared with other current state-of-the-art methods at specific proportions of labeled data, the best performance in semi-supervised segmentation was achieved by DVCPS, demonstrating that DVCPS can effectively address the extreme imbalance of labeled data.
- 9
- 8
- LSH-DETR: object detection algorithm for marine benthic organisms based on improved DETR8
The proposed LSH-DETR is more effective at locating small, dense marine benthic organisms and achieves a performance improvement of 13.8% MR when compared to the original DETR.
- Understanding the Generalization Performance of Spectral Clustering Algorithms8
This paper studies the excess risk bounds of the popular spectral clustering algorithms: relaxed RatioCut and relaxed NCut and proposes two novel algorithms that can penalize this quantity and, additionally, can cluster the out-of-sample data without re-eigendecomposition on the overall samples.
- 8
- 8
- 7
- 7
- 6
- 6
- Generative Model Watermarking Based on Human Visual System6
This paper introduces two HVS-based generative model watermarking methods, which are realized in RGB color space and YUV color space respectively and improves the fidelity of the model to be protected and has good universality compared with previous methods.
- Reversible data hiding in H.264/AVC videos based on hybrid-dimensional histogram modification6
A new reversible data hiding (RDH) scheme for H.264/AVC video based on hybrid histogram modification is proposed, which can achieve better performance in terms of bit rate variation and visual quality.
- 5
- 5
- 4
- UniFace++: Revisiting a Unified Framework for Face Reenactment and Swapping via 3D Priors4
This work delves into the intrinsic connections between the intrinsic connections between face reenactment and swapping from a more general training paradigm perspective, introducing a novel diffusion-based unified method UniFace++ and redefining both as target-oriented reconstruction tasks.
- 4
- Enhancing Grid-Based 3D Object Detection in Autonomous Driving With Improved Dimensionality Reduction4
A new point cloud backbone network for grid-based object detection, MDRNet, which is based on adaptive dimensionality reduction and multi-level spatial residual strategies, and designed to dynamically concentrate on the essential components of the object during 3D-to-BEV transformation.
- 4
- Text-driven Visual Prosody Generation for Embodied Conversational Agents4
This work proposes to take only transcripts as input to compute head movements, based on a statistical framework, and shows that the generated head animation is able to improve human perception in terms of naturalness and demonstrate that the headAnimation is synchronized with the input of synthetic speech.
- 3
- 3
- End-to-end image captioning based on reduced feature maps of deep learners pre-trained for object detection3
This research proposed an image caption generation method based on the most popular architecture with the YOLO as encoder and the LSTM as decoder for natural language sentences generation to describe the images.
- 2
- 2
- Partially Supervised Unpaired Multi-modal Learning for Label-Efficient Medical Image Segmentation2
This paper investigates the use of partially labeled data for label-efficient unpaired multi-modal learning and proposes a novel Decomposed partial class adaptation with snapshot Ensembled Self-Training (DEST) framework for it, which outperforms existing methods significantly.
- Matching strategy and skip-scale head configuration guideline based traffic object detection2
There is a rule of matching degrees between the object scale and the detection head across different input resolutions or detection scenes by careful analysis and this rule can handily and quantitatively guide the rational configuration of detection heads to effectively detect objects at vastly different scales.
- 2
- 2
- Generative Model Watermarking Suppressing High-Frequency Artifacts2
The main idea of the proposed framework is to design a new watermark embedding network that can suppress high-frequency artifacts by applying anti-aliasing, which indicates that the marked model not only maintains the performance very well on the original task, but also demonstrates better imperceptibility and robustness on the watermarking task.
- Exemplar-Based Image Colorization with A Learning Framework2
This paper proposes an automatic colorization method with a learning framework that decouples the colorization process and learning process so as to generate various color styles for the same gray image.
- 2
- 2
- 2
- 2
- 2
- Channel attention pyramid network for remote physiological measurement1
CAP-rPPG demonstrates outstanding robustness under various challenging conditions, such as varying lighting environments and physical motion, and consistently outperforms most state-of-the-art methods on both the UBFC-rPPG and PURE datasets.
- 1
- SiamRCSC: Robust siamese network with channel and spatial constraints for visual object tracking1
A high-order spatial constraint (HOSC) module is proposed, considering high-order correlations between features through recursively executing interactions in feature maps, and achieves competitive performance compared to SOTA trackers (CNN-based).
- 1
- FusionDepth: Complement Self-Supervised Monocular Depth Estimation with Cost Volume1
This paper proposes a multi-frame depth estimation framework which monocular depth can be refined continuously by multi- frame sequential constraints, leveraging a Bayesian fusion layer within several iterations.
- 1
- 1
- 1
- 1
- Correction to: Reversible data hiding in H.264/AVC videos based on hybrid-dimensional histogram modification1
A new reversible data hiding (RDH) scheme for H.264/AVC video based on hybrid histogram modification is proposed, which can achieve better performance in terms of bit rate variation and visual quality.
- Anti-recompression video watermarking algorithm without bit-rate increase1
A robust HEVC watermarking algorithm based on the analysis of the characteristics of HEVC recompression, which is robust to the recompression attack with constant quantization parameter and has good imperceptibility.
- 1
- Cooperative trajectory planning algorithm of USV-UAV with hull dynamic constraints1
A cooperative trajectory planning algorithm for the coupled USV-UAV system is proposed to ensure that USV can execute a safe and smooth path in the process of autonomous advance in multi-obstacle maps.
- 1
- 1
- –
- –
- –
- ERS: Enhancing representation supervision for contrastive-based semi-supervised medical image segmentation–
This work proposes a novel mutual learning framework to Enhance Representation Supervision (ERS) effectively and proposes a Representation-Logits Alignment module to align logit and representation spaces through prototype-guided pseudo-label generation and uncertainty-based calibration, enabling cross-space information exchange and collaborative supervision.
- Modelling of Nonlinear Oscillator System via Double Loop Radial Basis Function Neural Networks With Adaptive Radius and Lattices–
Helmholtz–Duffing and Vanderpol–Duffing are used as the testing objects of the nonlinear oscillator system while comparing with Deterministic Learning to show the effectiveness and superiority of the double loop RBFNN proposed.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Scatter Points in Space: 3D Detection from Multi-view Monocular Images–
This paper attempts to improve multi-view feature aggregation by proposing a learnable keypoints sampling method, which scatters pseudo surface points in 3D space, in order to keep data sparsity.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Hierarchical 3D surface reconstruction approach–
This paper proposes a hierarchical approach to 3D surface reconstruction based on Wavelet transformation of the 2D images, which can get the dense matching in different levels and shows that the algorithm is feasible.
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 2026-10-12. 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.