Research map / Detection and segmentation / Open-set and few-shot detection
iFAN: Image-Instance Full Alignment Networks for Adaptive Object Detection
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 iFAN instead of source-only Faster R-CNN | Faster R-CNN, VGG16 · SIM10K → Cityscapes | mAP: 34.9% → 46.9% | Helped |
| Use iFAN instead of source-only Faster R-CNN | Faster R-CNN, VGG16 · Cityscapes → Foggy Cityscapes | mAP: 16.9% → 35.3% | Helped |
| Add image-level alignment to source-only Faster R-CNN (ablation) | Faster R-CNN, VGG16 · SIM10K → Cityscapes | AP: 34.9 -> 43.0 | Helped |
| Add category-agnostic instance alignment to image-level alignment (ablation) | iFAN · SIM10K → Cityscapes | AP: 43.0 -> 46.1 | Helped |
| Add category-correlation instance alignment to image-level alignment (ablation) | iFAN · SIM10K → Cityscapes | AP: 43.0 -> 45.3 | Helped |
| Combine image-level, instance, and correlation alignments (ablation) | iFAN · SIM10K → Cityscapes | AP: 34.9 -> 46.9 | Helped |
| Add category-aware instance alignment rather than category-agnostic alignment (ablation) | iFAN · Cityscapes → Foggy Cityscapes | mAP: 33.1 -> 34.6 | Helped |
| Add category-correlation instance alignment rather than category-agnostic alignment (ablation) | iFAN · Cityscapes → Foggy Cityscapes | mAP: 33.1 -> 34.5 | Helped |
| Combine image-level, category-aware, and category-correlation alignments (ablation) | iFAN · Cityscapes → Foggy Cityscapes | mAP: 16.9 -> 35.3 | Helped |
| Align all four feature levels rather than one level (ablation) | iFAN image-level alignment · SIM10K → Cityscapes | AP: 41.3; 41.8; 39.4; 38.3 -> 43.0 | Helped |
| Start category-agnostic instance alignment at 10K rather than 30K iterations (ablation) | iFAN · SIM10K → Cityscapes | AP: 43.0% (image-level alignment only) vs 42.1% (start at 10K-th iters) | Hurt |
| Start category-agnostic instance alignment at 20K rather than 30K iterations (ablation) | iFAN · SIM10K → Cityscapes | AP: 43.6 vs 46.1 | Hurt |
Closest papers by meaning
- Adapting Object Detectors via Selective Cross-Domain Alignment (CVPR 2019 · similarity 0.93)
- Universal Domain Adaptive Object Detection via Dual Probabilistic Alignment (AAAI 2025 · similarity 0.89)
- Differential Alignment for Domain Adaptive Object Detection (AAAI 2025 · similarity 0.88)
- Mean Teacher DETR with Masked Feature Alignment: A Robust Domain Adaptive Detection Transformer Framework (AAAI 2024 · similarity 0.88)
- CAT: Exploiting Inter-Class Dynamics for Domain Adaptive Object Detection (CVPR 2024 · similarity 0.88)
- Large Self-Supervised Models Bridge the Gap in Domain Adaptive Object Detection (CVPR 2025 · 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.