Research map / Video / Video segmentation and tracking
Attribute-Driven Feature Disentangling and Temporal Aggregation for Video 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 attribute recognition loss to the softmax-only average-pooling baseline (ablation) | Video person re-identification model · MARS | mAP: 66.1 -> 68.2 | Helped |
| Add feature-level temporal weighting based on attribute confidence (ablation) | Attribute-supervised video person re-identification model · MARS | mAP: 68.2 -> 69.5 | Helped |
| Add semantic feature disentangling (ablation) | Attribute-supervised video person re-identification model · MARS | mAP: 68.2 -> 70.3 | Helped |
| Refine temporal weighting from feature level to sub-feature level (ablation) | Video person re-identification model with semantic feature groups · MARS | mAP: 69.5 -> 71.2 | Helped |
| Increase the re-weighting parameter from σ=0.5 to σ=1.2 (ablation) | Sub-feature temporal aggregation · MARS | mAP: 71.2 -> 70.5 | Hurt |
| Use weighted binary cross-entropy instead of binary cross-entropy for attribute transfer (ablation) | Attribute transfer model · MARS | mAP: 67.4 -> 69.2 | Helped |
| Add MMD regularization to the BCE-trained attribute transfer model (ablation) | Attribute transfer model · MARS | mAP: 67.4 -> 70.0 | Helped |
| Add MMD regularization to the WBCE-trained attribute transfer model (ablation) | Attribute transfer model · MARS | mAP: 69.2 -> 71.2 | Helped |
| Use attribute-driven feature disentangling and temporal aggregation | Video person re-identification model · MARS | mAP: 76.1 -> 78.2 | Helped |
| Use attribute-driven feature disentangling and temporal aggregation | Video person re-identification model · iLIDS-VID | Rank-1: 85.4 -> 86.3 | Helped |
| Use attribute-driven feature disentangling and temporal aggregation | Video person re-identification model · PRID-2011 | Rank-1: 93.2 -> 93.9 | Helped |
Closest papers by meaning
- Relation-Guided Spatial Attention and Temporal Refinement for Video-Based Person Re-Identification (AAAI 2020 · similarity 0.89)
- Rethinking Temporal Fusion for Video-Based Person Re-Identification on Semantic and Time Aspect (AAAI 2020 · similarity 0.88)
- Spatial and Temporal Mutual Promotion for Video-Based Person Re-Identification (AAAI 2019 · similarity 0.88)
- Video-Based Person Re-Identification via Self Paced Weighting (AAAI 2018 · similarity 0.87)
- Learning Heterogeneous Dictionary Pair with Feature Projection Matrix for Pedestrian Video Retrieval via Single Query Image (AAAI 2017 · similarity 0.87)
- Temporal Correlation Vision Transformer for Video Person Re-Identification (AAAI 2024 · similarity 0.86)
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.