Research map / Continual learning / Online and replay-based
TriRE: A Multi-Mechanism Learning Paradigm for Continual Knowledge Retention and Promotion
Preetha Vijayan, Prashant Shivaram Bhat, Bahram Zonooz, Elahe Arani
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 TriRE instead of the strongest rehearsal baseline | Seq-TinyImageNet Class-IL · Seq-TinyImageNet Class-IL | average top-1 accuracy (%): 20.14 vs 17.68 | Improved |
| Use TriRE instead of the strongest rehearsal baseline | Seq-TinyImageNet Task-IL · Seq-TinyImageNet Task-IL | average top-1 accuracy (%): 55.95 vs 52.60 | Improved |
| Use TriRE instead of CLS-ER | Seq-CIFAR10 Task-IL · Seq-CIFAR10 Task-IL | average top-1 accuracy (%): 93.59 vs 92.45 | Hurt |
| Use TriRE instead of CLS-ER | Seq-CIFAR100 Task-IL · Seq-CIFAR100 Task-IL | average top-1 accuracy (%): 72.01 vs 71.66 | Hurt |
| Add Revise to Retain + Rewind (ablation) | TriRE · Seq-CIFAR100 Task-IL | average top-1 accuracy (%): 72.09 -> 71.66 | Hurt |
| Add Revise to Retain + Rewind (ablation) | TriRE · Seq-CIFAR100 Class-IL | average top-1 accuracy (%): 43.03 -> 43.91 | Improved |
| Add Revise to Retain + Rewind (ablation) | TriRE · Seq-TinyImageNet Class-IL | average top-1 accuracy (%): 16.25 -> 20.14 | Improved |
| Add Revise to Retain + Rewind (ablation) | TriRE · Seq-TinyImageNet Task-IL | average top-1 accuracy (%): 48.89 -> 55.95 | Improved |
| Add Rewind to Retain (ablation) | TriRE · Seq-CIFAR100 Class-IL | average top-1 accuracy (%): 38.01 -> 43.03 | Improved |
| Add Rewind to Retain (ablation) | TriRE · Seq-TinyImageNet Class-IL | average top-1 accuracy (%): 11.54 -> 16.25 | Improved |
| Add Rewind to Retain (ablation) | TriRE · Seq-TinyImageNet Task-IL | average top-1 accuracy (%): 40.22 -> 48.89 | Improved |
| Add Revise to Retain (ablation) | TriRE · Seq-CIFAR100 Class-IL | average top-1 accuracy (%): 38.01 -> 33.08 | Hurt |
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.
- Foster — FOSTER: Feature Boosting and Compression for Class-Incremental Learning (ECCV 2022) · 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
- A Unified and General Framework for Continual Learning (ICLR 2024 · similarity 0.89)
- Thalamus: a brain-inspired algorithm for biologically-plausible continual learning and disentangled representations (ICLR 2023 · similarity 0.89)
- Unlocking the Power of Rehearsal in Continual Learning: A Theoretical Perspective (ICML 2025 · similarity 0.89)
- Learning without Isolation: Pathway Protection for Continual Learning (ICML 2025 · similarity 0.88)
- Do Your Best and Get Enough Rest for Continual Learning (CVPR 2025 · similarity 0.88)
- When to Learn and When to Stop: Quitting at the Optimal Time (Student Abstract) (AAAI 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.
