Research map / Video / Video segmentation and tracking
Appearance and Motion Enhancement for Video-Based Person Re-Identification
Read the paperCheck 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 AEM (attribute learning with pseudo labels) to I3D baseline (ablation) | AMEM (I3D) · MARS | mAP: 75.1 -> 78.4 | Helped |
| add MEM (c=3 future frame prediction) to I3D baseline (ablation) | AMEM (I3D) · MARS | mAP: 75.1 -> 79.0 | Helped |
| add MEM on top of AEM (ablation) | AMEM (I3D) · MARS | rank-1: 86.0 -> 86.7 | Helped |
| predict 3 frames instead of 1 in MEM (ablation) | AMEM (I3D) · MARS | rank-1: 85.4 -> 86.2 | Helped |
| predict 4 frames instead of 3 in MEM (ablation) | AMEM (I3D) · MARS | rank-1: 86.2 -> 85.6 | Hurt |
| use enhanced features only during training (drop branches at inference) (ablation) | AMEM (I3D) · MARS | rank-1: 83.3 -> 85.0 | Helped |
| remove texture MSE or adversarial loss from MEM (ablation) | AMEM (I3D) · MARS | rank-1 | No clear effect |
| add AMEM to R3D backbone (ablation) | R3D · MARS | rank-1: 56.1 -> 62.8 | Helped |
| add AMEM to P3D backbone (ablation) | P3D · MARS | rank-1: 68.8 -> 74.5 | Helped |
Closest papers by meaning
- Skeletons Speak Louder than Text: A Motion-Aware Pretraining Paradigm for Video-Based Person Re-Identification (AAAI 2026 · similarity 0.86)
- Spatial and Temporal Mutual Promotion for Video-Based Person Re-Identification (AAAI 2019 · similarity 0.85)
- Video-Based Person Re-Identification via Self Paced Weighting (AAAI 2018 · similarity 0.85)
- Rethinking Temporal Fusion for Video-Based Person Re-Identification on Semantic and Time Aspect (AAAI 2020 · similarity 0.85)
- Unleashing Potential of Unsupervised Pre-Training With Intra-Identity Regularization for Person Re-Identification (CVPR 2022 · similarity 0.85)
- SEAS: ShapE-Aligned Supervision for Person Re-Identification (CVPR 2024 · similarity 0.84)
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.