Research map / Continual learning / Class-incremental
A Theoretical Study on Solving Continual Learning
Gyuhak Kim, Changnan Xiao, Tatsuya Konishi, Zixuan Ke, Bing Liu
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 |
|---|---|---|---|
| replace ODIN with CSI as the per-task OOD detector (ablation) | HAT · C100-10T | CIL accuracy: 41.2 -> 63.3 | Improved |
| replace ODIN with CSI as the per-task OOD detector (ablation) | Sup · C100-20T | CIL accuracy: 36.4 -> 60.2 | Improved |
| replace ODIN with CSI as the per-task OOD detector (ablation) | HAT · C10-5T | CIL accuracy: 62.6 -> 87.8 | Improved |
| apply ODIN post-processing | DER++ · C100-10T | CIL accuracy: 53.71 -> 55.29 | Improved |
| apply ODIN post-processing | Sup · C100-10T | CIL accuracy: 44.58 -> 46.74 | Improved |
| apply ODIN post-processing | BiC · C100-10T | CIL accuracy: 52.92 -> 48.65 | Hurt |
| apply ODIN post-processing | PASS · C100-10T | CIL accuracy: 33.00 -> 31.00 | Hurt |
| apply ODIN post-processing | LwF · C100-10T | CIL accuracy: 45.26 -> 51.82 CIL while AUC 88.30 -> 87.11 | Mixed |
| add output calibration with small memory buffer (ablation) | HAT+CSI · C100-20T | CIL accuracy: 54.6 -> 58.0 | Improved |
| add output calibration with small memory buffer (ablation) | Sup+CSI · M-5T | CIL accuracy: 80.7 -> 81.0 | No clear effect |
| add CSI-based OOD detection to HAT | HAT · C100-10T TIL accuracy | TIL accuracy: 84.0 -> 92.0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- iCaRL — iCaRL: Incremental Classifier and Representation Learning (CVPR 2017) · better on 2 of 2 shared results
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
- Class Incremental Learning via Likelihood Ratio Based Task Prediction (ICLR 2024 · similarity 0.93)
- Continual Learning Using a Kernel-Based Method Over Foundation Models (AAAI 2025 · similarity 0.90)
- Make Domain Shift a Catastrophic Forgetting Alleviator in Class-Incremental Learning (AAAI 2025 · similarity 0.88)
- GACL: Exemplar-Free Generalized Analytic Continual Learning (NeurIPS 2024 · similarity 0.88)
- Learning Task-Aware Language-Image Representation for Class-Incremental Object Detection (AAAI 2024 · similarity 0.88)
- AnaCP: Toward Upper-Bound Continual Learning via Analytic Contrastive Projection (NeurIPS 2025 · 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.
