atlas

Research map / Detection and segmentation / CNN features

Differentiable Multi-Granularity Human Representation Learning for Instance-Aware Human Semantic Parsing

CVPR 2021 · Detection and segmentation › CNN features · approach: Joint detection and instance segmentation · 78 citations (OpenAlex)

Check 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 the complete multi-granularity parsing frameworkOurs · MHP_v2 valAP50^p: 25.1 -> 39.0Helped
Use the complete multi-granularity parsing frameworkOurs · DensePose-COCO testAP50^p: 37.6 -> 49.7Helped
Use the complete multi-granularity parsing frameworkOurs · PASCAL-Person-Part testAP50^r: 59.7 -> 59.0Mixed
Add coarse-to-fine DSPF estimation (ablation)DSPF baseline · MHP_v2 valAP50^p: 35.5 -> 36.9Helped
Add residual refinement to coarse-to-fine DSPF (ablation)Coarse-to-fine DSPF · MHP_v2 valAP50^p: 36.9 -> 38.1Helped
Add learnable feature warping to residual-refinement DSPF (ablation)Residual-refinement DSPF · MHP_v2 valAP50^p: 38.1 -> 38.9Helped
Replace differentiable matching with greedy matching (ablation)Human keypoint association · MHP_v2 valAP50^p: 39.0 -> 37.6Hurt
Replace differentiable matching with the other reported greedy algorithm (ablation)Human keypoint association · MHP_v2 valAP50^p: 39.0 -> 37.9Hurt

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