Research map / Detection and segmentation / Open-set and few-shot detection
StyleProto: Style-Augmented Prototype Learning for Cross-Domain Few-Shot Object Detection
Xi Yang, Quantao 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 |
|---|---|---|---|
| StyleProto over CD-ViTO | CD-ViTO · CD-FSOD 9 datasets avg (1-shot) | mAP: 11.8 -> 15.3 | Improved |
| StyleProto over CD-ViTO | CD-ViTO · CD-FSOD 9 datasets avg (5-shot) | mAP: 23.1 -> 25.5 | Improved |
| StyleProto over CD-ViTO | CD-ViTO · CD-FSOD 9 datasets avg (10-shot) | mAP: 26.7 -> 28.6 | Improved |
| StyleProto over CD-ViTO | CD-ViTO · COCO FSOD (10-shot) | nAP: 35.3 -> 35.5 | Improved |
| add all three modules (SGA, SPC, HPFA) (ablation) | CD-ViTO baseline · ArTaxOr (5-shot) | mAP: 47.9 -> 53.3 | Improved |
| add all three modules (SGA, SPC, HPFA) (ablation) | CD-ViTO baseline · UODD (5-shot) | mAP: 6.8 -> 8.4 | Improved |
| add all three modules (SGA, SPC, HPFA) (ablation) | CD-ViTO baseline · HRSID (5-shot) | mAP: 13.3 -> 16.2 | Improved |
| number of HPFA query prototypes per class beyond 5 (ablation) | StyleProto · CD-FSOD (5-shot/10-shot) | mAP: 43.8 -> 43.3 (5-shot, 5 to 7 queries) | Hurt |
| fusion coefficient alpha 0.3 vs high values (ablation) | StyleProto HPFA · CD-FSOD | mAP: peak at alpha=0.3 | Mixed |
Closest papers by meaning
- SCoA: Revisiting Domain Generalized Object Detection with Style-Conditioned Adaptation (ICML 2026 · similarity 0.89)
- Style Evolving along Chain-of-Thought for Unknown-Domain Object Detection (CVPR 2025 · similarity 0.88)
- Rethinking Prototype-based Similarity Learning for Few-Shot Object Detection (ECCV 2026 · similarity 0.88)
- Domain-RAG: Retrieval-Guided Compositional Image Generation for Cross-Domain Few-Shot Object Detection (NeurIPS 2025 · similarity 0.88)
- Fine-Grained Prototypes Distillation for Few-Shot Object Detection (AAAI 2024 · similarity 0.87)
- Multi-Perspective Data Augmentation for Few-shot Object Detection (ICLR 2025 · similarity 0.86)
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.
