AI Paper Atlas

Research map / 3D vision / Face and shape

To The Point: Correspondence-driven monocular 3D category reconstruction

Filippos Kokkinos, Iasonas Kokkinos

NeurIPS 2021 · 3D vision › Face and shape · approach: Articulated and non-rigid shape learning · 8 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 correspondence-driven pose and deformation optimization instead of the strongest reported baselineTTP on CUB with keypoint supervision · CUBPCK: 85.5 -> 93.4Improved
Use correspondence-driven reconstruction without keypoint supervision instead of ACSM with 8 camera hypothesesTTP on CUB · CUBmIoU: 0.693 -> 0.752Improved
Add keypoint supervisionTTP · PASCAL3D+ car3D mIoU: 0.665 -> 0.67No clear effect
Add texture loss (ablation)TTP with keypoint supervision · CUBmIoU: 0.752 -> 0.765Improved
Add texture loss (ablation)TTP without keypoint supervision · CUBmIoU: 0.667 -> 0.749Improved
Add visibility loss (ablation)TTP without keypoint supervision · CUBPCK: 9.3 -> 50.9Improved
Add equivariance loss (ablation)TTP without keypoint supervision · CUBPCK: 28.2 -> 50.9Improved
Use a deformable basis instead of a rigid template (ablation)TTP with keypoint supervision · CUBmIoU: 0.657 -> 0.765Improved
Use a deformable basis instead of a rigid template (ablation)TTP without keypoint supervision · CUBmIoU: 0.646 -> 0.749Improved
Increase basis size from 16 to 64 components (ablation)TTP with keypoint supervision · CUBmIoU: 0.765 -> 0.775Improved
Increase basis size from 16 to 64 components (ablation)TTP without keypoint supervision · CUBno clear effect: 0.749 -> 0.752No clear effect
Increase pose/deformation optimization iterations from 1 to 4 (ablation)TTP with keypoints · CUBmIoU: 0.732 -> 0.758Improved

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