Research map / Continual learning / Class-incremental
Dealing With Cross-Task Class Discrimination in Online Continual Learning
Yiduo Guo, Bing Liu, Dongyan Zhao
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 |
|---|---|---|---|
| Use GSA instead of the strongest reported baseline | Online class-incremental learning · MNIST, CIFAR10, CIFAR100, and TinyImageNet | accuracy: 91.4 vs 90.4; 64.6 vs 58.2; 49.7 vs 39.5; 33.2 vs 22.5 | Improved |
| remove the within-task classification loss for new classes (ablation) | GSA · CIFAR100; TinyImageNet | accuracy: 31.4 -> 30.1; 18.4 -> 17.5 | Hurt |
| remove the within-task classification loss for previous classes (ablation) | GSA · CIFAR100; TinyImageNet | accuracy: 31.4 -> 12.7; 18.4 -> 4.7 | Hurt |
| remove the cross-task classification loss (ablation) | GSA · CIFAR100; TinyImageNet | accuracy: 31.4 -> 29.9; 18.4 -> 17.2 | Hurt |
| remove the balanced mixed set Xmix (ablation) | GSA · CIFAR100; TinyImageNet | accuracy: 31.4 -> 30.1; 18.4 -> 17.5 | Hurt |
| replace GSA-CE with ordinary cross-entropy (ablation) | GSA · CIFAR100; TinyImageNet | accuracy: 31.4 -> 29.7; 18.4 -> 16.9 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- SSIL — SS-IL: Separated Softmax for Incremental Learning (ICCV 2021) · better on 2 of 2 shared results
- GDumb — GDumb: A Simple Approach that Questions Our Progress in Continual Learning (ECCV 2020) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Mitigating Catastrophic Forgetting in Online Continual Learning by Modeling Previous Task Interrelations via Pareto Optimization (ICML 2024 · similarity 0.89)
- Layerwise Proximal Replay: A Proximal Point Method for Online Continual Learning (ICML 2024 · similarity 0.89)
- Continual Learning Using a Kernel-Based Method Over Foundation Models (AAAI 2025 · similarity 0.88)
- Online Curvature-Aware Replay: Leveraging $\mathbf{2^{nd}}$ Order Information for Online Continual Learning (ICML 2025 · similarity 0.88)
- Defying Imbalanced Forgetting in Class Incremental Learning (AAAI 2024 · similarity 0.88)
- Non-exemplar Online Class-Incremental Continual Learning via Dual-Prototype Self-Augment and Refinement (AAAI 2024 · similarity 0.88)
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.
