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
- Qiang XuSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works69 from public data
- 3,320
- Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks1,364
This work introduces Grounded SAM, which uses Grounding DINO as an open-set object detector to combine with the segment anything model (SAM), which enables the detection and segmentation of any regions based on arbitrary text inputs and opens a door to connecting various vision models.
- Effective Whole-body Pose Estimation with Two-stages Distillation446
This work presents a two-stage pose Distillation for Whole-body Pose estimators, named DWPose, to improve their effectiveness and efficiency and releases a series of models with different sizes, from tiny to large, for satisfying various downstream tasks.
- FITS: Modeling Time Series with $10k$ Parameters376
FITS achieves performance comparable to state-of-the-art models for time series forecasting and anomaly detection tasks, while having a remarkably compact size of only approximately $10k$ parameters.
- Human Pose Regression with Residual Log-likelihood Estimation315
This work proposes a novel regression paradigm with Residual Log-likelihood Estimation (RLE) to capture the underlying output distribution and finds that for the first time, the proposed regression method is superior to the heatmap-based methods.
- SRNet: Improving Generalization in 3D Human Pose Estimation with a Split-and-Recombine Approach229
The proposed split-and-recombine approach, called SRNet, can be easily adapted to both single-image and temporal models, and it leads to appreciable improvements in the prediction of rare and unseen poses.
- Follow-Your-Emoji: Fine-Controllable and Expressive Freestyle Portrait Animation199
Follow-Your-Emoji, a diffusion-based framework for portrait animation, which animates a reference portrait with target landmark sequences, is presented and a facial fine-grained loss is proposed to improve the model’s ability of subtle expression perception and reference portrait appearance reconstruction.
- One-Stage 3D Whole-Body Mesh Recovery with Component Aware Transformer186
This work designs a one-stage pipeline for expressive whole-body mesh recovery, named OSX, without separate networks for each part, and designs a Component Aware Transformer (CAT) composed of a global body encoder and a local face/hand decoder.
- Learning Skeletal Graph Neural Networks for Hard 3D Pose Estimation170
This work proposes a hop-aware hierarchical channel-squeezing fusion layer to effectively extract relevant information from neighboring nodes while suppressing undesired noises in GNN learning and proposes a temporal-aware dynamic graph construction procedure that is robust and effective for 3D pose estimation.
- From Knowledge Distillation to Self-Knowledge Distillation: A Unified Approach with Normalized Loss and Customized Soft Labels166
This work unifies the formulations of the two tasks by decomposing and reorganizing the generic KD loss into a Normalized KD (NKD) loss and customized soft labels for both target class (image’s category) and non-target classes named Universal Self-KD (USKD).
- Direct Inversion: Boosting Diffusion-based Editing with 3 Lines of Code163
Direct Inversion, a novel technique achieving optimal performance of both branches with just three lines of code, is introduced, which not only yields superior performance across 8 editing methods but also achieves nearly an order of speed-up.
- HumanSD: A Native Skeleton-Guided Diffusion Model for Human Image Generation155
Experimental results show that HumanSD outperforms ControlNet in terms of pose control and image quality, particularly when the given skeleton guidance is sophisticated, particularly when the given skeleton guidance is sophisticated.
- Lite DETR : An Interleaved Multi-Scale Encoder for Efficient DETR154
Lite DETR is presented, a simple yet efficient end-to-end object detection framework that can effectively reduce the GFLOPs of the detection head by 60% while keeping 99% of the original performance, and the efficient encoder strategy can generalize well across existing DETR-based models.
- HumanTOMATO: Text-aligned Whole-body Motion Generation138
This work proposes a Text-aligned whOle-body Motion generATiOn framework, named HumanTOMATO, which is the first attempt to the authors' knowledge towards applicable holistic motion generation in this research area and has significant advantages in both the quality of generated motions and their alignment with text.
- SmoothNet: A Plug-and-Play Network for Refining Human Poses in Videos136
SmoothNet models the natural smoothness characteristics in body movements by learning the long-range temporal relations of every joint without considering the noisy correlations among joints, and improves the temporal smoothness of existing pose estimators significantly and enhances the estimation accuracy of those challenging frames as a side-effect.
- DiffSHEG: A Diffusion-Based Approach for Real-Time Speech-Driven Holistic 3D Expression and Gesture Generation104
This work proposes DiffSHEG, a Diffusion-based approach for Speech-driven Holistic 3D Expression and Gesture generation with arbitrary length, and introduces an outpainting-based sampling strategy for arbitrary long sequence generation in diffusion models, offering flexibility and computational efficiency.
- Human-Art: A Versatile Human-Centric Dataset Bridging Natural and Artificial Scenes99
The Human-Art dataset is introduced and contains 50k high-quality images with over 123k person instances from 5 natural and 15 artificial scenarios, which are annotated with bounding boxes, keypoints, self-contact points, and text information for humans represented in both 2D and 3D.
- 98
- Explicit Box Detection Unifies End-to-End Multi-Person Pose Estimation94
This paper presents a novel end-to-end framework with Explicit box Detection for multi-person Pose estimation, called ED-Pose, where it unifies the contextual learning between human-level and keypoint-level information and surpasses heatmap-based Top-down methods under the same backbone.
- 94
- GPAvatar: Generalizable and Precise Head Avatar from Image(s)89
A framework named GPAvatar is proposed that reconstructs 3D head avatars from one or several images in a single forward pass and introduces a dynamic point-based expression field driven by a point cloud to precisely and effectively capture expressions.
- DeepFuse: An IMU-Aware Network for Real-Time 3D Human Pose Estimation from Multi-View Image78
A two-stage fully 3D network to estimate human pose in 3D space by fusing body-worn Inertial Measurement Unit (IMU) data and multi-view images deeply and introduces an IMU-bone layer to fuse the IMU and vision data earlier at data level is proposed.
- IDOL: Instant Photorealistic 3D Human Creation from a Single Image71
A scalable feed-forward transformer model is developed to predict a 3D human Gaussian representation in a uniform space from a given human image, which demonstrates the ability to efficiently reconstruct photorealistic humans at 1K resolution from a single input image using a single GPU instantly.
- AiOS: All-in-One-Stage Expressive Human Pose and Shape Estimation62
A novel ali-in-one-stage framework, AiOS, for multiple expressive human pose and shape recovery without an additional human detection step is introduced, built upon DETR, which treats multi-person whole-body mesh recovery task as a progressive set prediction problem with various sequential detection.
- DeciWatch: A Simple Baseline for 10x Efficient 2D and 3D Pose Estimation61
DeciWatch introduces a simple yet effective sample-denoise-recover framework that only watches sparsely sampled frames, taking advantage of the continuity of human motions and the lightweight pose representation.
- 55
- Open-World Human-Object Interaction Detection via Multi-Modal Prompts51
MP-HOI is a powerful Multimodal Prompt-based HOI detector designed to leverage both textual descriptions for open-set generalization and visual exemplars for handling high ambiguity in descriptions, realizing HOI detection in the open world.
- ViTKD: Practical Guidelines for ViT feature knowledge distillation50
This paper designs a series of controlled experiments and derives three practical guidelines for ViT's feature distillation and proposes the feature-based method ViTKD which brings consistent and considerable improvement to the student.
- Follow-Your-Emoji-Faster: Towards Efficient, Fine-Controllable, and Expressive Freestyle Portrait Animation49
Stable Diffusion is enhanced with two key components: a expression-aware landmarks as explicit motion signals, which improve motion alignment, support exaggerated expressions, and reduce identity leakage; and a fine-grained facial loss that leverages both expression and facial masks to better capture subtle expressions and faithfully preserve the reference appearance.
- SMPLest-X: Ultimate Scaling for Expressive Human Pose and Shape Estimation49
This work investigates the impact of scaling up EHPS towards a family of generalist foundation models, and finetuning strategy turns the generalist into specialist models, allowing them to achieve further performance boosts.
- Motion-X++: A Large-Scale Multimodal 3D Whole-body Human Motion Dataset47
A scalable annotation pipeline is developed that can automatically capture 3D whole-body human motion and comprehensive textural labels from RGB videos and build the Motion-X dataset comprising 81.1K text-motion pairs, a large-scale multimodal 3D expressive whole-body human motion dataset.
- 44
- SkillMimic: Learning Basketball Interaction Skills from Demonstrations42
SkillMimic, a unified data-driven framework that fundamentally changes how agents learn interaction skills by eliminating the need for skill-specific rewards, is introduced, showing that a unified HOI imitation reward can effectively capture the essence of diverse interaction patterns from HOI datasets.
- Boosting Human-Object Interaction Detection with Text-to-Image Diffusion Model37
DiffHOI is introduced, a novel HOI detection scheme grounded on a pre-trained text-image diffusion model, which enhances the detector's performance via improved data diversity and HOI representation, and SynHOI, a class-balance, large-scale, and high-diversity synthetic dataset containing over 140K HOI images with fully triplet annotations.
- The Dawn of Video Generation: Preliminary Explorations with SORA-like Models35
Despite the emergence of DiT-based closed-source and open-source models, a comprehensive investigation into their capabilities and limitations remains lacking, and the rapid development has made it challenging for recent benchmarks to fully cover SORA-like models and recognize their significant advancements.
- MotionCraft: Crafting Whole-Body Motion with Plug-and-Play Multimodal Controls33
This paper proposes MotionCraft, a unified diffusion transformer that crafts whole-body motion with plug-and-play multimodal control with state-of-the-art performance on various standard motion generation tasks and introduces MC-Bench, the first available multimodal whole-body motion generation benchmark based on the unified SMPL-X format.
- VideoGen-Eval: Agent-based System for Video Generation Evaluation29
VideoGen-Eval is proposed, an agent evaluation system that integrates LLM-based content structuring, MLLM-based content judgment, and patch tools designed for temporal-dense dimensions to achieve a dynamic, flexible, and expandable video generation evaluation.
- 29
- X-Pose: Detecting Any Keypoints28
X-Pose is proposed, a novel end-to-end framework with multi-modal prompts to detect multi-object keypoints for any articulated, rigid, and soft objects within a given image, which demonstrates X-Pose's strong fine-grained keypoint localization and generalization abilities across image styles, object categories, and poses.
- detrex: Benchmarking Detection Transformers27
A unified, highly modular, and lightweight codebase called detrex is developed, which supports a majority of the mainstream DETR-based instance recognition algorithms, covering various fundamental tasks, including object detection, segmentation, and pose estimation.
- 24
- 22
- FrAug: Frequency Domain Augmentation for Time Series Forecasting22
This paper proposes simple yet effective frequency domain augmentation techniques that ensure the semantic consistency of augmented data-label pairs in forecasting, named FrAug, and shows that FrAug can boost the forecasting accuracy of TSF models in most cases.
- DPoser-X: Diffusion Model as Robust 3D Whole-Body Human Pose Prior19
This work introduces a Diffusion model as body Pose prior (DPoser) and extends it to DPoser-X for expressive whole-body human pose modeling, establishing a new benchmark for whole-body human pose prior modeling.
- 17
- FreeMan: Towards Benchmarking 3D Human Pose Estimation Under Real-World Conditions16
FreeMan, the first large-scale, multi-view dataset collected under the real-world conditions, is presented and evaluations of standard indoor/outdoor human sensing datasets reveal that FreeMan offers robust representation transferability in real and complex scenes.
- VFX Creator: Animated Visual Effect Generation with Controllable Diffusion Transformer15
This work proposes a novel paradigm for animated VFX generation as image animation, where dynamic effects are generated from user-friendly textual descriptions and static reference images, and introduces a specialized metric to evaluate the precision of temporal control.
- 15
- Information Bottleneck Approach to Spatial Attention Learning15
Extensive experiments show that the proposed IB-inspired spatial attention mechanism can yield attention maps that neatly highlight the regions of interest while suppressing backgrounds, and bootstrap standard DNN structures for visual recognition tasks (e.g., image classification, fine-grained recognition, cross-domain classification).
- 13
- 12
- LPM 1.0: Video-based Character Performance Model11
This work presents LPM 1.0 (Large Performance Model), a single-person full-duplex audio-visual conversational performance model, focusing on single-person full-duplex audio-visual conversational performance, and proposes LPM-Bench, the first benchmark for interactive character performance.
- 9
- 7
- 6
- 6
- 6
- 6
- A Comprehensive Benchmark for Neural Human Radiance Fields5
This work design and execute thorough studies into unified evaluation settings and metrics to establish a fair and reasonable benchmark for human NeRF models, and makes HumanNeRF from monocular videos generalizable, as the inaugural baseline.
- UniMC: Taming Diffusion Transformer for Unified Keypoint-Guided Multi-Class Image Generation3
A DiT-based framework, named UniMC, is designed to explore unifying controllable multi-class image generation, and a large-scale, high-quality, and diverse dataset designed for keypoint-guided human and animal image generation is proposed.
- 2
- 2
- –
- –
- –
- –
- –
- –
- –
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 2026-10-15. 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.