Research questionHow can surgical perception reconstruct 3D surfaces accurately from sparse endoscopic viewpoints?Clinical telerobots typically provide limited endoscopic viewpoints, leaving much of the surgical surface weakly observed. This makes it difficult to determine how reconstruction accuracy and coverage change as viewpoints are added, especially when tissue deforms. Latest papersRecent research connected to this question, newest first.MV-dVRK: A Multi-Viewpoint Benchmark for Spatial Surgical PerceptionThe source evaluates reconstruction using the ex-vivo MV-dVRK benchmark, which provides synchronized stereo views, surface geometry, camera poses, sparse-view test sets, and dynamic sequences spanning surgical tasks with increasing kinematic complexity and tissue deformation. It compares zero-shot monocular, stereo, multi-stereo, and multi-view approaches; the evidence covers benchmarked reconstruction performance rather than clinical deployment.research paper · Sep 2, 2026