Research map / Continual learning / Class-incremental
RBC: Rectifying the Biased Context in Continual Semantic Segmentation
Hanbin Zhao, Fengyu Yang, Xinghe Fu, Xi Li
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 context-rectified image duplet (original plus new-class-erased image) (ablation) | PLOP baseline · VOC-15-1 Overlapped | mIoU all classes: 54.64 -> 61.14 | Improved |
| add context-rectified image duplet (ablation) | PLOP baseline · VOC-15-1 Overlapped | mIoU new classes: 21.11 -> 31.06 | Improved |
| use erased duplet instead of copied original images (double) (ablation) | PLOP baseline · VOC-15-1 Overlapped | mIoU all classes: 48.73 -> 61.14 | Improved |
| add copied original images (double) to PLOP (ablation) | PLOP baseline · VOC-15-1 Overlapped | mIoU all classes: 54.64 -> 48.73 | Hurt |
| add biased-context-insensitive consistency loss (ablation) | Baseline+duplet · VOC-15-1 Overlapped | mIoU new classes: 31.06 -> 38.44 | Improved |
| add biased-context-insensitive consistency loss (ablation) | Baseline+duplet · VOC-15-1 Overlapped | mIoU all classes: 61.14 -> 62.14 | Improved |
| add biased-context-insensitive consistency loss (ablation) | Baseline+duplet · VOC-15-1 Overlapped | mIoU old classes (0-15): 70.54 -> 69.54 | Hurt |
| add adaptive class-balance strategy (ablation) | PLOP baseline · VOC-15-1 Overlapped | mIoU new classes: 21.11 -> 24.89 | Improved |
| add adaptive class-balance strategy (ablation) | PLOP baseline · VOC-15-1 Overlapped | mIoU old classes (0-15): 65.12 -> 65.35 | Improved |
| add adaptive class-balance strategy (ablation) | PLOP baseline · VOC-15-1 Overlapped | mIoU all classes: 54.64 -> 55.72 | Improved |
| replace PLOP with RBC full method | PLOP · VOC-15-1 Overlapped | mIoU all classes: 54.64 -> 62.14 | Improved |
| replace PLOP with RBC full method | PLOP · VOC-19-1 Overlapped | mIoU new classes: 37.35 -> 55.60 | Improved |
Closest papers by meaning
- A Survey on Continual Semantic Segmentation: Theory, Challenge, Method and Application (TPAMI 2024 · similarity 0.91)
- Beyond Background Shift: Rethinking Instance Replay in Continual Semantic Segmentation (CVPR 2025 · similarity 0.91)
- Preparing the Future for Continual Semantic Segmentation (ICCV 2023 · similarity 0.91)
- Decoupling Continual Semantic Segmentation (AAAI 2026 · similarity 0.90)
- Class Similarity Weighted Knowledge Distillation for Continual Semantic Segmentation (CVPR 2022 · similarity 0.90)
- TIKP: Text-to-Image Knowledge Preservation for Continual Semantic Segmentation (AAAI 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.
