Research map / Label-efficient and robust learning / Domain adaptation and generalization
CIA: Cluster-Instance Alignment for Unsupervised Day-Night Vehicle Re-Identification
Yongguo Ling, Chen Zhang, Yiming Liu, Wenhao Shao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 DGM cluster-level matching to the baseline (ablation) | Base (ADCA-style dual-stream) · DN-348 Day-to-Night | Rank-1 (%): 63.17 -> 65.32 | Improved |
| add MAA instance-level alignment to the baseline (ablation) | Base (ADCA-style dual-stream) · DN-348 Day-to-Night | Rank-1 (%): 63.17 -> 63.76 | Improved |
| add MAA to the baseline (ablation) | Base · DN-348 Night-to-Day | Rank-1 (%): 76.61 -> 78.22 | Improved |
| combine DGM and MAA (ablation) | Base · DN-348 Day-to-Night | Rank-1 (%): 65.32 -> 67.74 | Improved |
| replace PGM cluster matching with DGM (ablation) | Base · DN-348 Night-to-Day | Rank-1 (%): 77.52 -> 79.33 | Improved |
| replace MMD with MAA (ablation) | ADCA · DN-348 Day-to-Night | Rank-1 (%): 63.22 -> 64.32 | Improved |
| replace MarginMMD with MAA (ablation) | ADCA · DN-348 Night-to-Day | Rank-1 (%): 78.75 -> 78.88 | No clear effect |
| use continuous similarity-based dictionary scores instead of discrete scores (ablation) | DGM · DN-348 | Rank-1 (%): Day-to-Night 66.96 -> 67.74; Night-to-Day 80.36 -> 79.86 | Mixed |
| use continuous instead of fixed dictionary scores (ablation) | DGM · DN-348 Day-to-Night | Rank-1 (%): 66.92 -> 67.74 | Improved |
| apply CIA to pedestrian RegDB visible-thermal | Base · RegDB Visible-to-Thermal | Rank-1 (%): 67.20 -> 86.12 | Improved |
| CIA vs best prior method on DN-Wild day-to-night | CIA · DN-Wild Day-to-Night | Rank-1 (%): 48.63 -> 49.12 | Improved |
Closest papers by meaning
- Day-Night Cross-domain Vehicle Re-identification (CVPR 2024 · similarity 0.90)
- NightReID: A Large-Scale Nighttime Person Re-Identification Benchmark (AAAI 2025 · similarity 0.86)
- Multi-Spectral Vehicle Re-Identification: A Challenge (AAAI 2020 · similarity 0.85)
- Disentangled Feature Learning Network and a Comprehensive Benchmark for Vehicle Re-Identification (TPAMI 2021 · similarity 0.85)
- Alice Benchmarks: Connecting Real World Re-Identification with the Synthetic (ICLR 2024 · similarity 0.84)
- Infer the Whole from a Glimpse of a Part: Keypoint-Based Knowledge Graph for Vehicle Re-Identification (AAAI 2025 · similarity 0.83)
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.
