arXiv preprint / cs.RO

Estimate, Don't Imitate: Reusing Differentiable State-Based Policies for Visuomotor Control

作者：Denis Shcherba, Adrian Abel, Eckart Cobo-Briesewitz, Wojciech Samek, Marc Toussaint · 单位：BIFOLD – Berlin Institute for the Foundations of Learning and Data, Berlin, Germany., Robotics Institute Germany. · 会议/期刊：arXiv preprint · 方向：cs.RO · 发布日期：2026-09-29

2026-09-29🤖 操作与抓取
新标签打开 AlphaXiv
Estimate, Don't Imitate: Reusing Differentiable State-Based Policies for Visuomotor Control figure
AlphaXiv 中文概览（可滚动查看）