Research map / Vision-language and multimodal / Cross-modal learning
ProxyTTT: Proxy-driven Test-Time Training for Multi-modal Re-identification
Aihua Zheng, Zhaojun Liu, Xixi Wan, Chenglong Li, Jin Tang, Yan Yan
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 modality-common proxy (MCP) (ablation) | ProxyTTT baseline · RGBNT201 | mAP (%): 75.4 -> 77.0 | Improved |
| add modality-specific proxy (MSP) (ablation) | ProxyTTT · RGBNT201 | mAP (%): 77.0 -> 79.8 | Improved |
| add fine-grained reconstruction learning (FRL) (ablation) | ProxyTTT · RGBNT201 | mAP (%): 79.8 -> 82.3 | Improved |
| add PESA test-time training (ablation) | ProxyTTT · RGBNT201 | mAP (%): 82.3 -> 85.0 | Improved |
| add MPL to EDITOR (ablation) | EDITOR · RGBNT201 | mAP (%): 66.5 -> 69.3 | Improved |
| add MPL to DeMo (ablation) | DeMo · RGBNT201 | mAP (%): 77.4 -> 80.4 | Improved |
| use proxy-based pseudo-labels instead of classification head (ablation) | PESA · RGBNT201 | mAP (%): 82.6 -> 83.9 | Improved |
| add entropy-based selection to proxy pseudo-labels (ablation) | PESA · RGBNT201 | mAP (%): 83.9 -> 85.0 | Improved |
| add entropy selection to head pseudo-labels (ablation) | PESA with Head · RGBNT201 | Rank-1 (%): mAP 82.6 -> 83.0; R-1 86.4 -> 85.4 | Mixed |
| test with missing RGB and TIR modalities | ProxyTTT vs DeMo · RGBNT201 | Rank-1 (%): M(RGB+TIR) R-1 24.9 (DeMo) -> 24.2 | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- MambaPro — MambaPro: Multi-Modal Object Re-identification with Mamba Aggregation and Synergistic Prompt (AAAI 2025) · better on 4 of 4 shared results
- IDEA — IDEA: Inverted Text with Cooperative Deformable Aggregation for Multi-modal Object Re-Identification (CVPR 2025) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Heterogeneous Test-Time Training for Multi-Modal Person Re-identification (AAAI 2024 · similarity 0.89)
- MDReID: Modality-Decoupled Learning for Any-to-Any Multi-Modal Object Re-Identification (NeurIPS 2025 · similarity 0.88)
- Magic Tokens: Select Diverse Tokens for Multi-modal Object Re-Identification (CVPR 2024 · similarity 0.86)
- Multi-Modal Object Re-identification via Sparse Mixture-of-Experts (ICML 2025 · similarity 0.85)
- All in One Framework for Multimodal Re-identification in the Wild (CVPR 2024 · similarity 0.85)
- Alice Benchmarks: Connecting Real World Re-Identification with the Synthetic (ICLR 2024 · similarity 0.85)
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.
