Research map / Label-efficient and robust learning / Test-time adaptation
Continuous Adaptation for Interactive Object Segmentation by Learning from Corrections
Theodora Kontogianni, Michael Gygli, Jasper Uijlings, Vittorio Ferrari
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 image adaptation (IA) at test time | frozen model · Berkeley | mean clicks to reach target IoU: 5.4 -> 4.9 | Improved |
| add image sequence adaptation (SA) at test time | frozen model · YouTube-VOS | mean clicks to reach target IoU: 7.9 -> 6.9 | Improved |
| add image sequence adaptation (SA) at test time | frozen model · COCO unseen 6k | mean clicks to reach target IoU: 13.2 -> 10.0 | Improved |
| combine IA with SA | SA alone · COCO unseen 6k | mean clicks to reach target IoU: 10.0 -> 9.3 | Improved |
| add SA on top of IA | IA alone · Berkeley | mean clicks to reach target IoU: 4.9 -> 4.9 | No clear effect |
| add SA | frozen model · DRIONS-DB | mean clicks to reach target IoU: 13.3 -> 3.6 | Improved |
| add IA | frozen model · DRIONS-DB | mean clicks to reach target IoU: 13.3 -> 11.4 | Improved |
| combine IA with SA | SA alone · DRIONS-DB | mean clicks to reach target IoU: 3.6 -> 3.1 | Improved |
| combine IA with SA | SA alone · Rooftop Aerial | mean clicks to reach target IoU: 3.6 -> 3.6 | No clear effect |
| add SA (class specialization) | frozen model · COCO donut | mean clicks to reach target IoU: 11.6 -> 7.1 | Improved |
| add IA+SA (class specialization) | frozen model · COCO bench | mean clicks to reach target IoU: 15.1 -> 13.3 | Improved |
| train on user corrections in single image adaptation (ablation) | IA without corrections · COCO unseen 6k | mean clicks to reach target IoU: 13.2 -> 10.6 | Improved |
Closest papers by meaning
- Live Interactive Training for Video Segmentation (CVPR 2026 · similarity 0.88)
- Learning When and How to Update Memory for Video Object Segmentation (TPAMI 2026 · similarity 0.88)
- Large-Scale Interactive Object Segmentation With Human Annotators (CVPR 2019 · similarity 0.88)
- You Point, I Learn: Online Adaptation of Interactive Segmentation Models for Handling Distribution Shifts in Medical Imaging (ICLR 2026 · similarity 0.87)
- Test-Time Adaptation of Vision-Language Models for Open-Vocabulary Semantic Segmentation (NeurIPS 2025 · similarity 0.86)
- Rethinking Interactive Image Segmentation with Low Latency High Quality and Diverse Prompts (CVPR 2024 · 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.
