Zheqi Lv
3 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
VISUALTHINK-VLA introduces a visual intermediate-reasoning framework that guides action prediction using compact visual evidence, achieving high accuracy and significantly low latency for real-time Vision-Language-Action policies.
VISUALTHINK-VLA introduces a visual intermediate-reasoning framework that guides action prediction using compact visual evidence, achieving high accuracy and significantly low latency for real-time Vision-Language-Action policies.
The paper proposes Gazer, a framework that integrates multimodal large language model feedback into the AVM sampling loop for in-generation semantic correction.
Papers
Training-Free Semantic Correction for Autoregressive Visual Models
Junhao Chen, Chanyu Zhu, Zheqi Lv, Keting Yin +1 more
The paper proposes Gazer, a framework that integrates multimodal large language model feedback into the AVM sampling loop for in-generation semantic correction.