AI Paper Atlas

Research map / 3D vision / Human pose estimation

NIKI: Neural Inverse Kinematics With Invertible Neural Networks for 3D Human Pose and Shape Estimation

CVPR 2023 · 3D vision › Human pose estimation · approach: Transformer-based multi-view 3D pose · 75 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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 twist-and-swing mapping instead of one-stage mappingNIKI · 3DPW-XOCCPA-MPJPE: 64.4 -> 60.5Helped
Add the independence constraint to decouple pose and error (ablation)NIKI · 3DPW-XOCCMPJPE: 112.6 -> 110.7Helped
Add the independence constraint to decouple pose and error (ablation)NIKI · 3DPWMPJPE: 71.6 -> 71.3No clear effect
Add boundary constraints (ablation)NIKI · 3DPWMPJPE: 73.6 -> 71.3Helped
Add boundary constraints (ablation)NIKI · 3DPW-XOCCMPJPE: 112.5 -> 110.7Helped
Train in both forward and inverse directions (ablation)NIKI · 3DPWMPJPE: 74.0 -> 71.3Helped
Train in both forward and inverse directions (ablation)NIKI · 3DPW-XOCCMPJPE: 111.9 -> 110.7Helped
Add temporal affine coupling layersNIKI · 3DPW-XOCCMPJPE: 110.7 -> 88.9Helped
Add temporal affine coupling layersNIKI · 3DPWMPJPE: 71.3 -> 71.2Helped
Add heatmap conditioning (ablation)NIKI · 3DPWMPJPE: 71.3 -> 71.1Helped
Add heatmap conditioning (ablation)NIKI · 3DPW-XOCCMPJPE: 110.7 -> 110.8Hurt
Use NIKI instead of HybrIK3DPW-XOCC · 3DPW-XOCCPA-MPJPE: 98.7 -> 60.5Helped

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