Reproduce “SurgAM: Surgical Affordance Map Prediction with Multimodal Feature Fusion for Robot Autonomy” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Lei Song, Yonghao Long, Mengya Xu, Jiayi Geng, Xiuyuan Chen, Qi Dou
SurgAM predicts action-specific maps of candidate tissue regions for aspiration, clipping, and retraction, linking perception to an existing robot controller. It combines semantic and spatial visual features with surgical-context prompts and scene-guided decoding. It scores 1.642 NSS versus 1.446 for WorldAfford and succeeds in phantom retraction and clipping tasks in 50/50 and 49/50 trials, respectively, not live surgery.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.