AI Paper Atlas

Research map / Point clouds / Registration and matching

SOMA: Solving Optical Marker-Based MoCap Automatically

Nima Ghorbani, Michael J. Black

ICCV 2021 · Point clouds › Registration and matching · approach: Non-rigid point set registration · 46 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
add AMASS noise model to synthetic data (ablation)SOMA · HDM05accuracy: 94.73 -> 95.50Improved
add CAESAR bodies for body shape variation (ablation)SOMA · HDM05accuracy: 95.21 -> 95.50Improved
use Sinkhorn optimal transport instead of Log-Softmax (ablation)SOMA · HDM05accuracy: 91.51 -> 95.50Improved
add random marker placement noise (ablation)SOMA · HDM05accuracy: 89.41 -> 95.50Improved
add transformer self-attention layers (ablation)SOMA · HDM05accuracy: 11.36 -> 95.50Improved
use tracklet labeling instead of per-frame labelingSOMA · BMLmoviaccuracy: 95.90 -> 97.33Improved
use tracklet labeling instead of per-frame labelingSOMA · KITaccuracy: 94.97 -> 95.46Improved
train with full noise (occlusion and ghost) instead of base noise, tested on noisy dataSOMA · HDM05 (B+C+G test)accuracy: 73.95 -> 95.50Improved
train with full noise instead of base noise, tested on clean dataSOMA · HDM05 (B test)accuracy: 97.93 -> 95.74Hurt
train on varied AMASS bodies and motions (SOMA*) instead of dataset-specific synthetic markersSOMA · BMLrub (5 occlusions + ghost)accuracy: 95.13 -> 97.56Improved
train one model per reduced layout instead of a base model on the full layoutSOMA · HDM05 (3 markers removed)accuracy: 95.35 -> 91.78Hurt
train on real instead of synthetic markers (same motions)SOMA · BMLrub (0 occlusions)accuracy: 99.16 -> 99.08No clear effect

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