Research map / Image generation / GANs
Symbiotic Adversarial Learning for Attribute-based Person Search
Yu-Tong Cao, Jingya Wang, Dacheng Tao
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 single adversarial alignment to the embedding model (ablation) | Embed · PETA | mAP: 31.3 -> 35.0 | Improved |
| replace single adversarial learning with symbiotic adversarial learning (ablation) | Embed+adv · PETA | mAP: 35.0 -> 40.6 | Improved |
| add symbiotic adversarial learning to the embedding model (ablation) | Embed · PETA | rank1: 34.0 -> 44.0 | Improved |
| add sampling of unseen attribute combinations (ablation) | Embed+symb-adv · PETA | mAP: 40.6 -> 41.2 | Improved |
| add sampling of unseen attribute combinations (ablation) | Embed+symb-adv · PETA | rank1: 44.0 -> 47.0 | Improved |
| add augmented loss L_aug (use synthetic data in alignment) (ablation) | SAL · PETA | mAP: 35.4 -> 41.2 | Improved |
| add granularity-consistency loss L_consis (ablation) | SAL · PETA | mAP: 35.2 -> 41.2 | Improved |
| add granularity-consistency loss L_consis (ablation) | SAL · PETA | rank1: 39.5 -> 47.0 | Improved |
| use symbiotic iterative training instead of stage-wise training (ablation) | SAL · PETA | mAP: 35.0 -> 41.2 | Improved |
| use symbiotic iterative training instead of stage-wise training (ablation) | SAL · PETA | rank1: 41.0 -> 47.0 | Improved |
Closest papers by meaning
- ViTAA: Visual-Textual Attributes Alignment in Person Search by Natural Language (ECCV 2020 · similarity 0.84)
- ASMR: Learning Attribute-Based Person Search With Adaptive Semantic Margin Regularizer (ICCV 2021 · similarity 0.84)
- Person Search by Text Attribute Query As Zero-Shot Learning (ICCV 2019 · similarity 0.84)
- Joint Feature Synthesis and Embedding: Adversarial Cross-Modal Retrieval Revisited (TPAMI 2020 · similarity 0.84)
- OVL: One-View Learning for Human Retrieval (AAAI 2020 · similarity 0.84)
- Robust NAS under adversarial training: benchmark, theory, and beyond (ICLR 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.
