13 papers
Keep the Future, Drop the Rollout: RIFT for World Action Models
Chushan Zhang, Jinguang Tong, Xuesong Li +2
World action models (WAMs) condition robot actions on predicted futures, but iterative video rollout increases deployment latency. We ask whether action generation requires the evo…
Repurposing RGB-based Foundation Model for Depth Estimation on Thermal Images Using Hierarchical Supervision
Jie Hong, Tingtian Li, Xuesong Li +1
Depth estimation from thermal images is highly valuable for robotic applications in adverse conditions, such as nighttime and rainy weather. Recent studies have sought to transfer…
Structural Energy Guidance for View-Consistent Text-to-3D Generation
Qing Zhang, Jinguang Tong, Jing Zhang +2
Text-to-3D generation based on diffusion models often suffers from the Janus problem, leading to inconsistent geometry across viewpoints. This work identifies viewpoint bias in 2D…
EvoScene-VLA: Evolving Scene Beliefs Inside the Action Decoder for Chunked Robot Control
Chushan Zhang, Ruihan Lu, Jinguang Tong +3
Chunked vision-language-action (VLA) policies generate multi-step actions from one observation and typically re-observe after executing the action chunk. Earlier actions change the…
EgoSelf: From Memory to Personalized Egocentric Assistant
Yanshuo Wang, Yuan Xu, Xuesong Li +4
Egocentric assistants often rely on first-person view data to capture user behavior and context for personalized services. Since different users exhibit distinct habits, preference…
Adaptive and Balanced Re-initialization for Long-timescale Continual Test-time Domain Adaptation
Yanshuo Wang, Jinguang Tong, Jun Lan +5
Continual test-time domain adaptation (CTTA) aims to adjust models so that they can perform well over time across non-stationary environments. While previous methods have made cons…
Structural Energy-Guided Sampling for View-Consistent Text-to-3D
Qing Zhang, Jinguang Tong, Jie Hong +2
Text-to-3D generation often suffers from the Janus problem, where objects look correct from the front but collapse into duplicated or distorted geometry from other angles. We attri…
GS-2DGS: Geometrically Supervised 2DGS for Reflective Object Reconstruction
Jinguang Tong, Xuesong li, Fahira Afzal Maken +4
3D modeling of highly reflective objects remains challenging due to strong view-dependent appearances. While previous SDF-based methods can recover high-quality meshes, they are of…
Enhancing Features in Long-tailed Data Using Large Vision Model
Pengxiao Han, Changkun Ye, Jinguang Tong +4
Language-based foundation models, such as large language models (LLMs) or large vision-language models (LVLMs), have been widely studied in long-tailed recognition. However, the ne…
Maintain Plasticity in Long-timescale Continual Test-time Adaptation
Yanshuo Wang, Xuesong Li, Jinguang Tong +5
Continual test-time domain adaptation (CTTA) aims to adjust pre-trained source models to perform well over time across non-stationary target environments. While previous methods ha…
DGNS: Deformable Gaussian Splatting and Dynamic Neural Surface for Monocular Dynamic 3D Reconstruction
Xuesong Li, Jinguang Tong, Jie Hong +2
Dynamic scene reconstruction from monocular video is essential for real-world applications. We introduce DGNS, a hybrid framework integrating \underline{D}eformable \underline{G}au…
Improving Viewpoint Consistency in 3D Generation via Structure Feature and CLIP Guidance
Qing Zhang, Jinguang Tong, Jing Zhang +2
Despite recent advances in text-to-3D generation techniques, current methods often suffer from geometric inconsistencies, commonly referred to as the Janus Problem. This paper iden…
Latent-based Diffusion Model for Long-tailed Recognition
Pengxiao Han, Changkun Ye, Jieming Zhou +3
Long-tailed imbalance distribution is a common issue in practical computer vision applications. Previous works proposed methods to address this problem, which can be categorized in…