Interactive Visual Gym enables closed-loop evaluation and training of visual interaction models by jointly simulating task environments and human u...
论文概览
- arXiv ID:
2611.08286 - 发布机构 / 来源:Hugging Face Daily Papers
- 作者:Chenhao Zheng, Zixian Ma, Kai Li et al.
- 社区热度:13 Upvotes
核心摘要(Abstract / AI Summary)
Interactive Visual Gym enables closed-loop evaluation and training of visual interaction models by jointly simulating task environments and human users, exposing limitations in state-of-the-art VLMs and improving performance through closed-loop training.
