Research map / Detection and segmentation / Open-set and few-shot detection
Real-Time Source-Free Object Detection
Sairam V C Rebbapragada, Varun Gopal, Poornima Jain, Vineeth N Balasubramanian, Muhammad Haris Khan
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 Dual-Head Fusion (DHF) pseudo-labels (ablation) | MT (O2O), RT-SFOD-M · C2F | mAP: 45.2 -> 46.7 | Improved |
| add DHF (ablation) | MT (O2O), RT-SFOD-M · S2C | AP: 62.1 -> 65.1 | Improved |
| add DHF (ablation) | MT (O2O), RT-SFOD-M · K2C | AP: 50.6 -> 53.0 | Improved |
| add MARD loss (ablation) | MT (O2O), RT-SFOD-M · C2F | mAP: 45.2 -> 46.3 | Improved |
| add MARD loss (ablation) | MT (O2O), RT-SFOD-M · S2C | AP: 62.1 -> 65.3 | Improved |
| add MARD on top of DHF (ablation) | MT (O2O) + DHF · C2F | mAP: 46.7 -> 47.4 | Improved |
| add mean-teacher self-training with O2O labels over AdaBN (ablation) | AdaBN · C2F | mAP: 37.4 -> 45.2 | Improved |
| use O2O instead of O2M pseudo-labels in mean teacher (ablation) | MT · C2F | mAP: 44.7 -> 45.2 | Improved |
| add AdaBN before self-training (ablation) | source-trained model · C2F | mAP: 30.5 -> 37.4 | Improved |
| add variance term of MARD (ablation) | MT (O2O) + DHF · C2F | mAP: 46.7 -> 47.2 | Improved |
| add covariance term of MARD (ablation) | MT (O2O) + DHF · C2F | mAP: 46.7 -> 47.0 | Improved |
| use DHF instead of directly merging O2O with O2M+NMS (ablation) | MT · C2F | mAP: 45.5 -> 46.7 | Improved |
Closest papers by meaning
- Simplifying Source-Free Domain Adaptation for Object Detection: Effective Self-Training Strategies and Performance Insights (ECCV 2024 · similarity 0.92)
- Beyond Boundaries: Leveraging Vision Foundation Models for Source-Free Object Detection (AAAI 2026 · similarity 0.91)
- Enhancing Source-Free Domain Adaptive Object Detection with Low-confidence Pseudo Label Distillation (ECCV 2024 · similarity 0.89)
- Dynamic Retraining-Updating Mean Teacher for Source-Free Object Detection (ECCV 2024 · similarity 0.89)
- YOLOv10: Real-Time End-to-End Object Detection (NeurIPS 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.
