Research map / Video / Video segmentation and tracking
Learning Multi-Granular Hypergraphs 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 multi-granular hypergraph learning to the frame-pooled baseline (ablation) | ResNet50 video re-ID · MARS | top-1 accuracy: 85.7 -> 88.7 | Helped |
| Add multi-granular hypergraph learning to the frame-pooled baseline (ablation) | ResNet50 video re-ID · iLIDS-VID | top-1 accuracy: 79.7 -> 84.3 | Helped |
| Add multi-granular hypergraph learning to the frame-pooled baseline (ablation) | ResNet50 video re-ID · PRID-2011 | top-1 accuracy: 88.6 -> 93.8 | Helped |
| Add node-level attention for graph-level feature aggregation (ablation) | HGNN · MARS | top-1 accuracy: 88.7 -> 89.2 | Helped |
| Add node-level attention for graph-level feature aggregation (ablation) | HGNN · iLIDS-VID | top-1 accuracy: 84.3 -> 85.1 | Helped |
| Add node-level attention for graph-level feature aggregation (ablation) | HGNN · PRID-2011 | top-1 accuracy: 93.8 -> 94.2 | Helped |
| Add mutual information loss between graph features (ablation) | HGNN + node attention · MARS | top-1 accuracy: 89.2 -> 90.0 | Helped |
| Add mutual information loss between graph features (ablation) | HGNN + node attention · MARS | mAP: 84.8 -> 85.8 | Helped |
| Add mutual information loss between graph features (ablation) | HGNN + node attention · iLIDS-VID | top-1 accuracy: 85.1 -> 85.6 | Helped |
| Add mutual information loss between graph features (ablation) | HGNN + node attention · PRID-2011 | top-1 accuracy: 94.2 -> 94.8 | Helped |
| Add temporal granularities 1 and 3 to range 1 (ablation) | MGH with spatial granularity 1 · MARS | mAP: 82.3 -> 83.2 | Helped |
| Add temporal range 5 to temporal granularities 1 and 3 (ablation) | MGH with spatial granularity 1 · MARS | mAP: 83.2 -> 83.3 | Helped |
Closest papers by meaning
- Multi-Granularity Reference-Aided Attentive Feature Aggregation for Video-Based Person Re-Identification (CVPR 2020 · similarity 0.88)
- Keypoint Message Passing for Video-Based Person Re-identification (AAAI 2022 · similarity 0.88)
- Spatial-Temporal Graph Convolutional Network for Video-Based Person Re-Identification (CVPR 2020 · similarity 0.88)
- Spatial-Temporal Correlation and Topology Learning for Person Re-Identification in Videos (CVPR 2021 · similarity 0.86)
- Diversity Regularized Spatiotemporal Attention for Video-Based Person Re-Identification (CVPR 2018 · similarity 0.86)
- Unsupervised Domain Adaptation for Person Re-identification via Heterogeneous Graph Alignment (AAAI 2021 · similarity 0.85)
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.