Research map / Detection and segmentation / Open-set and few-shot detection
Less Is Better: Sparse Instance Learning for Cross-Domain Few-Shot Object Detection
Yali Huang, Jie Mei, Ziyi Wu, Yiming Yang, Hongru Zhao, Mingyuan Jiu, Hichem Sahbi
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 query sparsity to CD-ViTO (ablation) | CD-ViTO · CD-FSOD avg (5-shot) | mAP: 26.1 -> 26.3 | Improved |
| add instance sparsity to CD-ViTO (ablation) | CD-ViTO · CD-FSOD avg (5-shot) | mAP: 26.1 -> 26.2 | Improved |
| combine instance and query sparsity (ablation) | CD-ViTO · CD-FSOD avg (5-shot) | mAP: 26.1 -> 27.3 | Improved |
| add query sparsity to CD-ViTO (ablation) | CD-ViTO · Clipart1k (1-shot) | mAP: 17.7 -> 20.4 | Improved |
| add instance sparsity on top of query sparsity (ablation) | CD-ViTO+MQS · Clipart1k (1-shot) | mAP: 20.4 -> 23.2 | Improved |
| add prototype calibration (InfoNCE) loss (ablation) | CD-ViTO+MQS+MIS · Clipart1k (1-shot) | mAP: 23.2 -> 24.5 | Improved |
| add query sparsity to CD-ViTO (ablation) | CD-ViTO · NEU-DET (1-shot) | mAP: 3.6 -> 3.3 | No clear effect |
| raise instance sparsity threshold from 0.2 to 0.9 (ablation) | SI-ViTO · Clipart1k (1-shot) | AP: 24.3 -> 23.3 | Hurt |
| use moderate instance sparsity (zeta 0.2) vs none (ablation) | SI-ViTO · Clipart1k (1-shot) | AP: 23.6 -> 24.3 | Improved |
| replace CD-ViTO with SI-ViTO | CD-ViTO · CD-FSOD avg (1-shot) | mAP: 13.9 -> 16.3 | Improved |
| replace CD-ViTO with SI-ViTO | CD-ViTO · CD-FSOD avg (10-shot) | mAP: 29.6 -> 30.4 | Improved |
| replace CD-ViTO with SI-ViTO | CD-ViTO · UODD (10-shot) | mAP: 7.0 -> 7.0 | No clear effect |
Closest papers by meaning
- Cross-Domain Few-Shot Object Detection via Enhanced Open-Set Object Detector (ECCV 2024 · similarity 0.92)
- AcroFOD: An Adaptive Method for Cross-Domain Few-Shot Object Detection (ECCV 2022 · similarity 0.90)
- Fine-Grained Prototypes Distillation for Few-Shot Object Detection (AAAI 2024 · similarity 0.89)
- Domain-RAG: Retrieval-Guided Compositional Image Generation for Cross-Domain Few-Shot Object Detection (NeurIPS 2025 · similarity 0.88)
- Weakly Supervised Few-Shot Object Detection with DETR (AAAI 2024 · similarity 0.87)
- Few-Shot Object Detection with Foundation Models (CVPR 2024 · similarity 0.87)
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.
