Diffusion Stabilizer Policy for Automated Surgical Robot Manipulations
Chonlam Ho, Jianshu Hu, Lei Song, Hesheng Wang, Qi Dou, Yutong Ban · 2025
Read Diffusion Stabilizer Policy for Automated Surgical Robot Manipulations by Chonlam Ho, Jianshu Hu, Lei Song, Hesheng Wang, Qi Dou, Yutong Ban (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-01252.txt.
Description
Chonlam Ho and co-authors proposed a learning method in 2025 for robots that assist in surgery. Such robots could make operations more precise, but teaching them is hard because good demonstrations are scarce and failed tries are usually thrown away.
Their method first trains a stabilizer on clean demonstrations only. It then keeps learning from a mix of clean and disturbed runs, keeping updates whose action errors look sound. The design is based on diffusion models, which build actions step by step. In tests across several surgical settings, the method beat rivals when demonstrations were clean. It also held up better when demonstrations were disturbed, a common real world problem.
Catalog record
| CREATOR | Chonlam Ho, Jianshu Hu, Lei Song, Hesheng Wang, Qi Dou, Yutong Ban |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 39 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-01252