AI Paper Atlas

Research map / 3D vision / Human pose estimation

Learning To Know Where To See: A Visibility-Aware Approach for Occluded Person Re-Identification

Jinrui Yang, Jiawei Zhang, Fufu Yu, Xinyang Jiang, Mengdan Zhang, Xing Sun, Ying-Cong Chen, Wei-Shi Zheng

ICCV 2021 · 3D vision › Human pose estimation · approach: Multi-person pose estimation · 100 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
add region visibility discriminator (VRD) (ablation)PCB baseline with VGTri · Occluded-DukeRank-1: 47.0 -> 62.2Improved
add region visibility discriminator (VRD) (ablation)PCB baseline · Occluded-REIDRank-1: 71.8 -> 80.6Improved
add visibility-guided hard triplet loss (VGTri) (ablation)PCB baseline · Occluded-DukeRank-1: 41.7 -> 47.0Improved
add visibility-guided hard triplet loss (VGTri) (ablation)PCB with VRD · Occluded-DukeRank-1: 61.3 -> 62.2Improved
VGTri instead of traditional hard triplet loss (ablation)Ours · Occluded-DukeRank-1: 61.2 -> 62.2Improved
learned soft visibility scores instead of hard pose-detector part labels (ablation)Ours · Occluded-DukeRank-1: 48.1 -> 62.2Improved
use manual ground-truth part labels instead of learned visibility (ablation)Ours · Partial-REIDRank-1: 85.7 -> 88.0No clear effect
reduce keypoints from 18 to 6 (ablation)Ours · Occluded-DukeRank-1: 62.2 -> 60.7No clear effect
reduce keypoints from 18 to 6 (ablation)Ours · Occluded-REIDRank-1: 81.0 -> 77.7Hurt
reduce keypoints from 14 to 6 (ablation)HOReID · Occluded-DukeRank-1: 55.1 -> 49.2Hurt
use manual ground-truth pose cuesPGFA · Occluded-DukeRank-1: 51.9 -> 52.8Improved
use Mask R-CNN instead of AlphaPose keypoints (ablation)Ours · Occluded-DukeRank-1: 62.2 -> 62.9No clear effect

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