Research map / Continual learning / Class-incremental
Learning With Style: Continual Semantic Segmentation Across Tasks and Domains
Marco Toldo, Umberto Michieli, Pietro Zanuttigh
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 L_kd^o distillation loss (ablation) | LwS · CS->BDD->IDD step 2, CS | mIoU: 43.04 -> 44.17 | Improved |
| add L_kd^o distillation loss (ablation) | LwS · BDD->IDD->CS step 2, BDD | mIoU: 32.33 -> 34.03 | Improved |
| add L_kd^o distillation loss (ablation) | LwS · IDD->CS->BDD step 2, IDD | mIoU: 36.13 -> 36.60 | Improved |
| add L_kd^o distillation loss (ablation) | LwS · CS->BDD->IDD step 1, CS | mIoU: 50.77 -> 53.31 | Improved |
| add L_kd^o distillation loss (ablation) | LwS · IDD->CS->BDD step 1, IDD | mIoU: 43.30 -> 43.95 | No clear effect |
| add L_kd^o distillation loss (ablation) | LwS · BDD->IDD->CS step 1, BDD | mIoU: 54.34 -> 54.53 (current), 41.36 -> 43.98 (past) | Hurt |
| use LwS instead of best class-incremental competitor UCD | LwS · IDD->CS->BDD step 1, IDD | mIoU: 26.45 -> 43.95 | Improved |
| use LwS instead of UCD | LwS · CS->BDD->IDD step 2, CS | mIoU: 37.34 -> 44.17 | Improved |
| use LwS instead of UCD | LwS · Mapillary generalization, CS->BDD->IDD step 2 | mIoU: 34.07 -> 40.43 | Improved |
| use LwS instead of UCD | LwS · Mapillary generalization, IDD->CS->BDD step 2 | mIoU: 25.89 -> 39.16 | Improved |
| add self-stylization to fine-tuning | fine-tuning · CS->BDD->IDD step 2, CS | mIoU: 12.10 -> 13.68 | Improved |
| add self-stylization to fine-tuning | fine-tuning · BDD->IDD->CS step 2, BDD | mIoU: 5.45 -> 14.24 | Improved |
Closest papers by meaning
- Modeling the Background for Incremental Learning in Semantic Segmentation (CVPR 2020 · similarity 0.89)
- A Survey on Continual Semantic Segmentation: Theory, Challenge, Method and Application (TPAMI 2024 · similarity 0.88)
- Learning at a Glance: Towards Interpretable Data-Limited Continual Semantic Segmentation via Semantic-Invariance Modelling (TPAMI 2024 · similarity 0.88)
- Advancing Cross-domain Discriminability in Continual Learning of Vision-Language Models (NeurIPS 2024 · similarity 0.87)
- Towards Continual Universal Segmentation (CVPR 2025 · similarity 0.86)
- Beyond Background Shift: Rethinking Instance Replay in Continual Semantic Segmentation (CVPR 2025 · 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.
