AI Paper Atlas

Research map / Point clouds / Shape and surface completion

CaSPR: Learning Canonical Spatiotemporal Point Cloud Representations

Davis Rempe, Tolga Birdal, Yongheng Zhao, Zan Gojcic, Srinath Sridhar, Leonidas Guibas

NeurIPS 2020 · Point clouds › Shape and surface completion · approach: Unsupervised shape correspondence · 36 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 TPointNet++ instead of PointNet++ w/ TimeT-NOCS canonicalization · CarsSpatial Err: 0.0510 -> 0.0118Improved
Use CaSPR instead of PointFlowSequence reconstruction · Cars, 10 observed framesEMD: 12.838 -> 10.103Improved
Use CaSPR instead of PointFlowSequence reconstruction · Chairs, 10 observed framesEMD: 17.267 -> 13.009Improved
Use CaSPR instead of PointFlowSequence reconstruction · Airplanes, 10 observed framesEMD: 9.500 -> 6.026Improved
Use CaSPR rather than OFlowWarping Cars reconstruction · 10 observed timestepsCD: 1.512 -> 0.955Improved
Reverse the input sequence timestampsCaSPR · Held-out rigid car sequencesEMD: 7.114 -> 88.938Hurt
Add canonicalization loss (ablation)CaSPR · Rigid cars, 10 observed framesCD: 0.605 -> 0.566Improved
Add canonicalization loss (ablation)CaSPR · Rigid cars, 10 observed framesEMD: 11.482 -> 10.103Improved
Add static/dynamic latent factorization (ablation)CaSPR · Rigid cars, 10 observed framesCD: 0.635 -> 0.566Improved
Add static/dynamic latent factorization (ablation)CaSPR · Rigid cars, 10 observed framesNFE: 101.3 -> 39.6Improved
Add pairwise input augmentation (ablation)CaSPR · Rigid cars, 10 observed framesCD: 0.577 -> 0.566Improved
Add pairwise input augmentation (ablation)CaSPR · Rigid cars, 10 observed framesEMD: 10.253 -> 10.103Improved

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