AI Paper Atlas

Research map / Video / Video segmentation and tracking

Pyramid Spatial-Temporal Aggregation for Video-Based Person Re-Identification

Yingquan Wang, Pingping Zhang, Shang Gao, Xia Geng, Hu Lu, Dong Wang

ICCV 2021 · Video › Video segmentation and tracking · approach: Spatial-temporal video person re-identification · 132 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 pyramid aggregation structure (ablation)baseline · MARSRank-1 (%): 88.8 -> 90.6Improved
add pyramid aggregation structure (ablation)baseline · DukeMTMC-VIDRank-1 (%): 95.9 -> 97.3Improved
add spatial reference attention (SRA) on pyramid (ablation)pyramid structure · MARSRank-1 (%): 90.6 -> 90.9Improved
add temporal reference attention (TRA) on pyramid (ablation)pyramid structure · MARSRank-1 (%): 90.6 -> 91.2Improved
use both SRA and TRA with pyramid (ablation)pyramid structure · MARSRank-1 (%): 90.6 -> 91.5Improved
use pyramid instead of adjacent reference aggregation (ARAS) (ablation)STAM · MARSRank-1 (%): 90.2 -> 91.5Improved
use pyramid instead of global reference aggregation (GRAS) (ablation)STAM · MARSRank-1 (%): 90.1 -> 91.5Improved
use pyramid instead of adjacent reference aggregation (ARAS) (ablation)STAM · DukeMTMC-VIDRank-1 (%): 96.8 -> 98.3Improved
add residual block in STAM (ablation)PSTA · MARSRank-1 (%): 91.3 -> 91.5No clear effect
increase number of STAM stages from 0 to 3 (ablation)PSTA · MARSRank-1 (%): 88.8 -> 91.5Improved
increase number of STAM stages from 0 to 3 (ablation)PSTA · DukeMTMC-VIDRank-1 (%): 96.9 -> 98.3Improved
PSTA instead of AP3Dvideo ReID backbone · MARSRank-1 (%): 90.7 -> 91.5Improved

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