Research map / Continual learning / Class-incremental
A Model or 603 Exemplars: Towards Memory-Efficient Class-Incremental Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use Memo's shared generalized blocks and task-specific specialized blocks | CIL models · CIFAR100 | AUC-A: 10.74 -> 10.85 | Helped |
| Use Memo's shared generalized blocks and task-specific specialized blocks | CIL models · ImageNet100 | AUC-A: 699.0 -> 713.0 | Helped |
| Use Memo's memory-efficient expandable architecture | CIFAR100 Base0 Inc5 CIL · CIFAR100 Base0 Inc5 | average accuracy: 69.52 -> 73.16 | Helped |
| Use Memo's memory-efficient expandable architecture | CIFAR100 Base0 Inc10 CIL · CIFAR100 Base0 Inc10 | average accuracy: 71.47 -> 72.37 | Helped |
| Add extra exemplars instead of storing more backbones | CIL methods at small memory budgets · CIFAR100 and ImageNet100 | accuracy | Mixed |
| Freeze specialized blocks from previous tasks (ablation) | Memo · CIFAR100 | incremental accuracy | Helped |
| Keep shared generalized blocks trainable rather than freezing them (ablation) | Memo with limited base classes · CIFAR100 Base10 | incremental accuracy | Helped |
| Freeze shared generalized blocks (ablation) | Memo with many base classes · CIFAR100 Base50 | incremental accuracy | Helped |
Closest papers by meaning
- Long-Tail Class Incremental Learning via Independent Sub-prototype Construction (CVPR 2024 · similarity 0.90)
- eTag: Class-Incremental Learning via Embedding Distillation and Task-Oriented Generation (AAAI 2024 · similarity 0.90)
- Gradient Reweighting: Towards Imbalanced Class-Incremental Learning (CVPR 2024 · similarity 0.89)
- Harnessing Neural Unit Dynamics for Effective and Scalable Class-Incremental Learning (ICML 2024 · similarity 0.89)
- Defying Imbalanced Forgetting in Class Incremental Learning (AAAI 2024 · similarity 0.89)
- Knowledge Memorization and Rumination for Pre-trained Model-based Class-Incremental Learning (CVPR 2025 · similarity 0.89)
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.