244 citations · 439 across the 46 of their papers we have counts for
41 papers · 1 filter
Structured Hyperedge Adaptation for Parameter-Efficient Fine-Tuning of Vision Transformers
Edwin Kwadwo Tenagyei, Lei Wang, Ugochukwu Ejike Akpudo +2
Parameter-efficient fine-tuning (PEFT) has become a practical solution for adapting large pretrained vision transformers (ViTs) to downstream tasks while updating only a small subs…
Test-Time Self-Adaptive Conditioning for Stable Audio-Driven Talking-Head Generation
Zhicheng Zhang, Lei Wang, Yu Zhang +1
Audio-driven talking-head generation has achieved remarkable progress with recent models such as AniTalker, FLOAT, and Sonic. Despite their success, most existing approaches rely o…
Trust-Aware Joint Feature-Prediction Discrepancy for Robust Domain Adaptation
Xi Ding, Lei Wang, Syuan-Hao Li +1
Domain adaptation aims to mitigate performance degradation caused by distribution shifts between a labeled source domain and an unlabeled or sparsely labeled target domain. Most ex…
Uncertainty-DTW for Sequences and Visual Tokens
Lei Wang, Syuan-Hao Li, Yongsheng Gao +1
Aligning structured data is a fundamental problem in computer vision and machine learning, underlying tasks such as time series analysis, human action recognition, and visual repre…
Guided Path Sampling: Steering Diffusion Models Back on Track with Principled Path Guidance
Haosen Li, Wenshuo Chen, Shaofeng Liang +3
Iterative refinement methods based on a denoising-inversion cycle are powerful tools for enhancing the quality and control of diffusion models. However, their effectiveness is crit…
POLARIS: Projection-Orthogonal Least Squares for Robust and Adaptive Inversion in Diffusion Models
Wenshuo Chen, Haosen Li, Shaofeng Liang +6
The Inversion-Denoising Paradigm, which is based on diffusion models, excels in diverse image editing and restoration tasks. We revisit its mechanism and reveal a critical, overloo…
Learning Time in Static Classifiers
Xi Ding, Lei Wang, Piotr Koniusz +1
Real-world visual data rarely presents as isolated, static instances. Instead, it often evolves gradually over time through variations in pose, lighting, object state, or scene con…
LUMA: Low-Dimension Unified Motion Alignment with Dual-Path Anchoring for Text-to-Motion Diffusion Model
Haozhe Jia, Wenshuo Chen, Yuqi Lin +8
While current diffusion-based models, typically built on U-Net architectures, have shown promising results on the text-to-motion generation task, they still suffer from semantic mi…
Video Understanding by Design: How Datasets Shape Video Models
Lei Wang, Syuan-Hao Li, Piotr Koniusz +1
Research in video understanding has advanced rapidly, driven by increasingly diverse datasets and more powerful model architectures. While existing surveys typically organize progr…
Puzzles: Unbounded Video-Depth Augmentation for Scalable End-to-End 3D Reconstruction
Jiahao Ma, Lei Wang, Miaomiao liu +2
Multi-view 3D reconstruction remains a core challenge in computer vision. Recent methods, such as DUST3R and its successors, directly regress pointmaps from image pairs without rel…
Feature Hallucination for Self-supervised Action Recognition
Lei Wang, Piotr Koniusz
Understanding human actions in videos requires more than raw pixel analysis; it relies on high-level semantic reasoning and effective integration of multimodal features. We propose…
ANT: Adaptive Neural Temporal-Aware Text-to-Motion Model
Wenshuo Chen, Kuimou Yu, Haozhe Jia +8
While diffusion models advance text-to-motion generation, their static semantic conditioning ignores temporal-frequency demands: early denoising requires structural semantics for m…
SRLoRA: Subspace Recomposition in Low-Rank Adaptation via Importance-Based Fusion and Reinitialization
Haodong Yang, Lei Wang, Md Zakir Hossain
Low-Rank Adaptation (LoRA) is a widely adopted parameter-efficient fine-tuning (PEFT) method that injects two trainable low-rank matrices (A and B) into frozen pretrained models. W…
Evolving Skeletons: Motion Dynamics in Action Recognition
Jushang Qiu, Lei Wang
Skeleton-based action recognition has gained significant attention for its ability to efficiently represent spatiotemporal information in a lightweight format. Most existing approa…
Physics-Informed Representation Alignment for Sparse Radio-Map Reconstruction
Haozhe Jia, Wenshuo Chen, Zhihui Huang +7
Radio map reconstruction is essential for enabling advanced applications, yet challenges such as complex signal propagation and sparse observational data hinder accurate reconstruc…
Free-T2M: Robust Text-to-Motion Generation for Humanoid Robots via Frequency-Domain
Wenshuo Chen, Haozhe Jia, Songning Lai +5
Enabling humanoid robots to synthesize complex, physically coherent motions from natural language commands is a cornerstone of autonomous robotics and human-robot interaction. Whil…
Quo Vadis, Anomaly Detection? LLMs and VLMs in the Spotlight
Xi Ding, Lei Wang
Video anomaly detection (VAD) has witnessed significant advancements through the integration of large language models (LLMs) and vision-language models (VLMs), addressing critical…
Do Language Models Understand Time?
Xi Ding, Lei Wang
Large language models (LLMs) have revolutionized video-based computer vision applications, including action recognition, anomaly detection, and video summarization. Videos inherent…
When Spatial meets Temporal in Action Recognition
Huilin Chen, Lei Wang, Yifan Chen +2
Video action recognition has made significant strides, but challenges remain in effectively using both spatial and temporal information. While existing methods often focus on eithe…
Learnable Expansion of Graph Operators for Multi-Modal Feature Fusion
Dexuan Ding, Lei Wang, Liyun Zhu +2
In computer vision tasks, features often come from diverse representations, domains (e.g., indoor and outdoor), and modalities (e.g., text, images, and videos). Effectively fusing…
TrackNetV4: Enhancing Fast Sports Object Tracking with Motion Attention Maps
Arjun Raj, Lei Wang, Tom Gedeon
Accurately detecting and tracking high-speed, small objects, such as balls in sports videos, is challenging due to factors like motion blur and occlusion. Although recent deep lear…
Motion meets Attention: Video Motion Prompts
Qixiang Chen, Lei Wang, Piotr Koniusz +1
Videos contain rich spatio-temporal information. Traditional methods for extracting motion, used in tasks such as action recognition, often rely on visual contents rather than prec…
SATO: Stable Text-to-Motion Framework
Wenshuo Chen, Hongru Xiao, Erhang Zhang +4
Is the Text to Motion model robust? Recent advancements in Text to Motion models primarily stem from more accurate predictions of specific actions. However, the text modality typic…
Optimizing Calibration by Gaining Aware of Prediction Correctness
Yuchi Liu, Lei Wang, Yuli Zou +2
Model calibration aims to align confidence with prediction correctness. The Cross-Entropy (CE) loss is widely used for calibrator training, which enforces the model to increase con…