AI Paper Atlas

Research map / 3D vision / Human pose estimation

Revitalizing Optimization for 3D Human Pose and Shape Estimation: A Sparse Constrained Formulation

Taosha Fan, Kalyan Vasudev Alwala, Donglai Xiang, Weipeng Xu, Todd Murphey, Mustafa Mukadam

ICCV 2021 · 3D vision › Human pose estimation · approach: Transformer-based multi-view 3D pose · 25 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 sparse constrained formulation instead of dense unconstrained optimizationSMPL pose and shape optimization · Human3.6Moptimization time (s): 20 -> 0.004Improved
use sparse constrained formulation in place of dense one (ablation)Gauss-Newton direction, SMPL and SMPL+H · CPU time Experiments 1-3CPU timeImproved
increase number of measurements N from 120 to 600 (ablation)sparse constrained method · SMPL / SMPL+H CPU timespeedupImproved
use SMPL+H (51 joints) instead of SMPL (23 joints) (ablation)sparse constrained method · SMPL / SMPL+H CPU timespeedup: CPU time ratio 4.49 -> 1Improved
increase number of shape parameters P from 0 to 10 (ablation)sparse constrained method · SMPL / SMPL+H CPU timespeedupHurt
use SCOPE instead of regression method SPINHuman3.6M Protocol 1 · Human3.6M Protocol 1MPJPE (mm): 65.6 -> 61.5Improved
use SCOPE instead of regression method SPINHuman3.6M Protocol 1 · Human3.6M Protocol 1PA-MPJPE (mm): 44.6 -> 48.2Hurt
use SCOPE instead of SPINMPI-INF-3DHP rigid aligned · MPI-INF-3DHP (rigid aligned)MPJPE (mm): 67.5 -> 62.1Improved
use SCOPE instead of SPIN3DPW · 3DPWMPJPE (mm): 96.9 -> 98.6Hurt
use SCOPE instead of XNect optimization3DPW · 3DPWMPJPE (mm): 134.2 -> 98.6Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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