Research map / Label-efficient and robust learning / Recognition and long-tail
VIGA: View-Conditioned and Identity-Guided Adaptation for Aerial-Ground Person Re-Identification
XU ZHANG, Feng Yining, ZUYU ZHANG
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 VC-LoRA (ablation) | VIGA · LAGPeR A->G | Rank-1: 41.79 -> 42.78 | Improved |
| add VC-LoRA (ablation) | VIGA · AG-ReID.v2 A->C | mAP: 80.84 -> 82.01 | Improved |
| add VC-LoRA (ablation) | VIGA · AG-ReID.v2 A->W | Rank-1: 91.44 -> 90.86 | Hurt |
| add VC-LoRA (ablation) | VIGA · LAGPeR G->A+G | Rank-1: 24.39 -> 24.39 | No clear effect |
| add IGS on top of VC-LoRA (ablation) | VIGA · LAGPeR G->A+G | Rank-1: 24.39 -> 25.11 | Improved |
| add IGS on top of VC-LoRA (ablation) | VIGA · AG-ReID.v2 A->W | Rank-1: 90.86 -> 90.36 | Hurt |
| soft masking instead of hard masking (ablation) | VIGA · LAGPeR A->G | Rank-1: 41.76 -> 43.47 | Improved |
| soft masking instead of hard masking (ablation) | VIGA · LAGPeR G->A+G | Rank-1: 25.25 -> 25.11 | Mixed |
| retain all tokens (rho=100%) instead of 87.5% (ablation) | VIGA · LAGPeR A->G | Rank-1: 43.47 -> 42.78 | Hurt |
| use rank 16 instead of rank 8 (ablation) | VIGA · LAGPeR A->G | Rank-1: 41.76 -> 43.47 | Improved |
| use rank 16 instead of rank 32 (ablation) | VIGA · LAGPeR A->G | Rank-1: 42.02 -> 43.47 | Improved |
| VIGA vs AG-ReIDv1 (uses attributes) on wearable | VIGA · AG-ReID.v2 A->W | Rank-1: 93.67 -> 90.36 | Hurt |
Closest papers by meaning
- View-decoupled Transformer for Person Re-identification under Aerial-ground Camera Network (CVPR 2024 · similarity 0.90)
- SeCap: Self-Calibrating and Adaptive Prompts for Cross-view Person Re-Identification in Aerial-Ground Networks (CVPR 2025 · similarity 0.90)
- AG-VPReID: A Challenging Large-Scale Benchmark for Aerial-Ground Video-based Person Re-Identification (CVPR 2025 · similarity 0.90)
- GSAlign: Geometric and Semantic Alignment Network for Aerial-Ground Person Re-Identification (NeurIPS 2025 · similarity 0.89)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.84)
- Feature-Level Adversarial Attacks and Ranking Disruption for Visible-Infrared Person Re-identification (NeurIPS 2024 · 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.
