Research map / Video / Video segmentation and tracking
Spatial-Temporal Graph Convolutional Network for Video-Based Person Re-Identification
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add temporal GCN modeling of patch relations across frames (ablation) | global-branch Re-ID model · MARS | mAP: 80.76 -> 81.97 | Helped |
| add spatial GCN modeling of within-frame patch relations (ablation) | global-branch Re-ID model · MARS | mAP: 80.76 -> 82.17 | Helped |
| jointly optimize spatial and temporal GCN branches (ablation) | STGCN · MARS | mAP: 82.17 -> 83.70 | Helped |
| combine the separate spatial and temporal branch models by score sum instead of jointly optimizing them (ablation) | STGCN · MARS | rank-1: 89.95 -> 89.55 | Hurt |
| replace temporal GCN layers with fully connected layers (ablation) | STGCN · MARS | mAP: 83.70 -> 81.51 | Hurt |
| replace spatial GCN layers with fully connected layers (ablation) | STGCN · MARS | mAP: 83.70 -> 82.65 | Hurt |
| replace both GCN branches with fully connected layers (ablation) | STGCN · MARS | rank-1: 89.95 -> 88.94 | Hurt |
| use four TGCN layers instead of three (ablation) | STGCN · MARS | rank-1: best Rank-1 is 90.35% when M = 4 | Helped |
| use three TGCN layers instead of four (ablation) | STGCN · MARS | mAP: best mAP is 83.70% then M = 3 | Helped |
| use two SGCN layers (ablation) | STGCN · MARS | performance | Helped |
| use four patches per frame rather than two or eight in TGCN (ablation) | STGCN · MARS | performance | Helped |
| use four patches per frame rather than two or eight in SGCN (ablation) | STGCN · MARS | performance | Helped |
Closest papers by meaning
- Spatial-Temporal Correlation and Topology Learning for Person Re-Identification in Videos (CVPR 2021 · similarity 0.93)
- See the Forest for the Trees: Joint Spatial and Temporal Recurrent Neural Networks for Video-Based Person Re-Identification (CVPR 2017 · similarity 0.92)
- Temporal-Enhanced Convolutional Network for Person Re-Identification (AAAI 2018 · similarity 0.91)
- Keypoint Message Passing for Video-Based Person Re-identification (AAAI 2022 · similarity 0.91)
- Relation-Guided Spatial Attention and Temporal Refinement for Video-Based Person Re-Identification (AAAI 2020 · similarity 0.90)
- Spatial and Temporal Mutual Promotion for Video-Based Person Re-Identification (AAAI 2019 · similarity 0.90)
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.