Research map / Video / Video segmentation and tracking
PhD Learning: Learning With Pompeiu-Hausdorff Distances for Video-Based Vehicle 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 |
|---|---|---|---|
| use PhD set-to-set metric learning with ID loss instead of triplet plus ID loss | Video re-ID · MARS | mAP: 82.2 -> 83.6 | Helped |
| use PhD set-to-set metric learning with ID loss instead of triplet plus ID loss | Video re-ID · PRID2011 | R1: 87.6 -> 92.1 | Helped |
| use PhD rather than batch-hard triplet metric learning | Video re-ID · MARS with distractor noise | mAP: 70.0 -> 80.4 | Helped |
| use PhD with k = S/2 instead of vanilla PhD with k = 1 (ablation) | Video re-ID · MARS, S = 6 | mAP: 82.7 -> 83.6 | Helped |
| use PhD with k = S/2 instead of vanilla PhD with k = 1 (ablation) | Video re-ID · PRID2011, S = 6 | R1: 92.1 -> 88.8 | Hurt |
| increase tracklet length from S = 4 to S = 8 (ablation) | PhD video re-ID · VVeRI-901 | mAP: 39.7 -> 42.0 | Helped |
| increase tracklet length from S = 6 to S = 8 (ablation) | PhD video re-ID · VVeRI-901 | mAP: 41.8 -> 42.0 | No clear effect |
| use random rather than even frame sampling (ablation) | PhD video re-ID · VVeRI-901 | mAP: 41.8 -> 45.7 | Helped |
| use random consecutive-sequence sampling rather than even frame sampling (ablation) | PhD video re-ID · VVeRI-901 | mAP: 41.8 -> 45.2 | Helped |
| use average temporal pooling instead of max pooling (ablation) | Video re-ID · VVeRI-901 | mAP: 40.5 -> 41.8 | Helped |
| use average pooling instead of attention aggregation (ablation) | Video re-ID · VVeRI-901 | mAP: 39.8 -> 41.8 | Helped |
| use average pooling instead of LSTM aggregation (ablation) | Video re-ID · VVeRI-901 | mAP: 37.2 -> 41.8 | Helped |
Closest papers by meaning
- Learning Coarse-to-Fine Structured Feature Embedding for Vehicle Re-Identification (AAAI 2018 · similarity 0.88)
- Deep Relative Distance Learning: Tell the Difference Between Similar Vehicles (CVPR 2016 · similarity 0.86)
- Alice Benchmarks: Connecting Real World Re-Identification with the Synthetic (ICLR 2024 · similarity 0.85)
- Part-Regularized Near-Duplicate Vehicle Re-Identification (CVPR 2019 · similarity 0.85)
- Infer the Whole from a Glimpse of a Part: Keypoint-Based Knowledge Graph for Vehicle Re-Identification (AAAI 2025 · similarity 0.83)
- CLIP-driven View-aware Prompt Learning for Unsupervised Vehicle Re-identification (AAAI 2025 · similarity 0.82)
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.