8 papers
Exposure Bias Can Alleviate Itself via Directional and Frequency Rectification in Flow Matching
Guanbo Huang, Jingjia Mao, Fanding Huang +9
Flow Matching (FM) has achieved remarkable generative performance, yet it suffers from exposure bias due to discrepancies between training and inference. Existing mitigation strate…
OSDTW: Optimal Shared Depth and Task Weighting for Long-Tailed Recognition
Chang Chu, Qingyue Zhang, Shao-Lun Huang +1
Long-tailed recognition suffers from a persistent head--tail trade-off: improving tail performance often degrades head accuracy and can increase training instability. Despite stron…
Sparse Compositional Flow Matching by geometric assembly from motion primitives
Yan Tang, Yuanbo Tang, Tingyu Cao +2
Embodied trajectories, such as the executable motion sequences of robotic manipulators, underwater vehicles, and mobile robots, are a fundamental output of embodied AI. Modern gene…
CAVE: A Structured Credit Assignment Approach for Fragmented Visual Evidence Reasoning
Tengda Guo, Jie Leng, Hanlei Li +6
Vision-Language Models (VLMs) have achieved strong performance on general multimodal reasoning, yet remain challenged in integrating nonlocal visual information to support semantic…
ReflexFlow: Rethinking Learning Objective for Exposure Bias Alleviation in Flow Matching
Guanbo Huang, Jingjia Mao, Fanding Huang +8
Despite tremendous recent progress, Flow Matching methods still suffer from exposure bias due to discrepancies in training and inference. This paper investigates the root causes of…
VFEM: Visual Feature Empowered Multivariate Time Series Forecasting with Cross-Modal Fusion
Yanlong Wang, Hang Yu, Jian Xu +7
Large time series foundation models often adopt channel-independent architectures to handle varying data dimensions, but this design ignores crucial cross-channel dependencies. Mea…
A High-Dimensional Statistical Method for Optimizing Transfer Quantities in Multi-Source Transfer Learning
Qingyue Zhang, Haohao Fu, Guanbo Huang +7
Multi-source transfer learning provides an effective solution to data scarcity in real-world supervised learning scenarios by leveraging multiple source tasks. In this field, exist…
RCA-NOC: Relative Contrastive Alignment for Novel Object Captioning
Jiashuo Fan, Yaoyuan Liang, Leyao Liu +2
In this paper, we introduce a novel approach to novel object captioning which employs relative contrastive learning to learn visual and semantic alignment. Our approach maximizes c…