Research map / Continual learning / Class-incremental
Dynamic Residual Classifier for Class Incremental Learning
Xiuwei Chen, Xiaobin Chang
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 DRC to MAF pipeline (ablation) | MAF · ImageNet100 B0 10 steps | Avg accuracy: 72.01 -> 78.65 | Improved |
| add DRC to MEC pipeline | MEC · ImageNet100 B0 10 steps | Avg accuracy: 70.08 -> 71.41 | Improved |
| add DRC to MDT pipeline | MDT · CIFAR100 B0 10 steps | Avg accuracy: 67.69 -> 70.80 | Improved |
| add logit adjustment (LA) to MAF+DRC (ablation) | MAFDRC · CIFAR100 B0 10 steps | Avg accuracy: 72.80 -> 73.97 | Improved |
| merge branch layers (DRC) instead of keeping all branches (RC) (ablation) | MAF residual classifier · ImageNet100 B0 10 steps | Avg accuracy: 79.26 -> 79.66 | No clear effect |
| use DRC instead of weight aligning for imbalance (ablation) | MAF · ImageNet100 B0 10 steps | Avg accuracy: 73.94 -> 78.65 | Improved |
| add DRC to MAF under LT-CIL | MAF · ImageNet100 Ordered LT-CIL 10 steps | Avg accuracy: 35.83 -> 59.65 | Improved |
| use MAFDRC instead of DER w/o P on CIFAR100 | CIFAR100 B0 10 steps · CIFAR100 B0 10 steps | Avg accuracy: 75.71 -> 73.97 | Hurt |
| use MAFDRC instead of FOSTER B4 | ImageNet100 B0 10 steps · ImageNet100 B0 10 steps | Avg accuracy: 76.54 -> 79.66 | Improved |
Closest papers by meaning
- Prediction Error-based Classification for Class-Incremental Learning (ICLR 2024 · similarity 0.86)
- Gradient Reweighting: Towards Imbalanced Class-Incremental Learning (CVPR 2024 · similarity 0.86)
- Long-Tail Class Incremental Learning via Independent Sub-prototype Construction (CVPR 2024 · similarity 0.85)
- Defying Imbalanced Forgetting in Class Incremental Learning (AAAI 2024 · similarity 0.84)
- Towards Macro-AUC Oriented Imbalanced Multi-Label Continual Learning (AAAI 2025 · similarity 0.84)
- Rebalancing Multi-Label Class-Incremental Learning (AAAI 2025 · similarity 0.84)
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.
