Research map / Architectures and efficiency / Attention and transformers
Uncertainty Modeling with Second-Order Transformer for Group Re-identification
Quan Zhang, Jian-Huang Lai, Zhanxiang Feng, Xiaohua Xie
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 |
|---|---|---|---|
| use SOT instead of MACG | G-ReID · CSG | Rank-1: 63.2 -> 91.7 | Improved |
| use SOT instead of MACG | G-ReID · DukeGroup | Rank-1: 57.4 -> 72.7 | Improved |
| use SOT instead of MACG | G-ReID · RoadGroup | Rank-1: 84.5 -> 86.4 | Improved |
| add MVM and LVM (ablation) | SOT certainty-modeling baseline · DukeGroup | Rank-1: 65.91 -> 72.73 | Improved |
| add MVM and LVM (ablation) | SOT certainty-modeling baseline · DukeGroup | mAP: 75.00 -> 78.90 | Improved |
| add MVM (ablation) | SOT with LVM · DukeGroup | Rank-1: 68.18 -> 72.73 | Improved |
| add LVM (ablation) | SOT with MVM · DukeGroup | Rank-1: 67.05 -> 72.73 | Improved |
| replace the proposed probability distribution with fixed p = 0.1, fixed p = 0.4, or a uniform distribution (ablation) | SOT member variation modeling · DukeGroup | Rank-1 | Hurt |
| replace the group feature transformer with GNN (ablation) | SOT inter-member module · DukeGroup | Rank-1 | Hurt |
| replace normalized layout embedding with random sequential embedding (ablation) | SOT layout modeling · DukeGroup | Rank-1: 68.18 vs 72.73 | Hurt |
| remove layout embedding (ablation) | SOT layout modeling · DukeGroup | Rank-1: 67.05 vs 72.73 | Hurt |
Closest papers by meaning
- Uncertainty Modeling for Group Re-Identification (IJCV 2024 · similarity 0.92)
- Unsupervised Group Re-identification via Adaptive Clustering-Driven Progressive Learning (AAAI 2024 · similarity 0.87)
- Transformer Learns Optimal Variable Selection in Group-Sparse Classification (ICLR 2025 · similarity 0.83)
- Exploring the Camera Bias of Person Re-identification (ICLR 2025 · similarity 0.83)
- UGG-ReID: Uncertainty-Guided Graph Model for Multi-Modal Object Re-Identification (NeurIPS 2025 · similarity 0.82)
- View-decoupled Transformer for Person Re-identification under Aerial-ground Camera Network (CVPR 2024 · 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.
