AI Paper Atlas

Research map / 3D vision / 3D generation

SHands: A Multi-View Dataset and Benchmark for Surgical Hand-Gesture and Error Recognition Toward Medical Training

Le Ma, Thiago Freitas dos Santos, Nadia Magnenat-Thalmann, Katarzyna Wac

CVPR 2026 · 3D vision › 3D generation · approach: 2D-to-3D pretrained representations · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use transformer backbone (TimeSformer) instead of CNN (SlowFast)video backbone · SHands single-view gesture recognitionTop-1 accuracy (%): 55.7 -> 58.4Improved
use masked pretraining (VideoMAE-L) instead of VideoMAE-BVideoMAE · SHands single-view gesture recognitionTop-1 accuracy (%): 63.4 -> 65.9Improved
use InternVideo2 instead of VideoMAE-Lvideo backbone · SHands single-view gesture recognitionTop-1 accuracy (%): 65.9 -> 68.9Improved
use five views (DVANet) instead of single-view InternVideo2gesture recognition · SHands gesture recognitionTop-1 accuracy (%): 68.9 -> 73.9Improved
use DVANet instead of MVAction in multi-viewmulti-view gesture recognition · SHands multi-view gesture recognitionMacro-F1: 0.704 -> 0.736Improved
use multi-view ViewCon instead of single-view VideoMAE-Berror detection · SHands error detection (trainee subset)Top-1 accuracy: 60.4 -> 66.2Improved
use DVANet with all five views instead of single-view VideoMAE-Berror detection · SHands error detection (trainee subset)Top-1 accuracy: 60.4 -> 68.5Improved
test on unseen views C4-C5 instead of seen viewsViewCon · SHands cross-view generalizationTop-1 accuracy (%): 68.4 -> 63.2Hurt
test on unseen views C4-C5 instead of seen viewsDVANet · SHands cross-view generalizationTop-1 accuracy (%): 72.8 -> 68.5Hurt
use DVANet instead of ViewCon on unseen viewscross-view generalization · SHands cross-view generalization (unseen views)Top-1 accuracy (%): 63.2 -> 68.5Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy