Research map / Continual learning / Online and replay-based
Training Networks in Null Space of Feature Covariance for Continual Learning
Shipeng Wang, Xiaorong Li, Jian Sun, Zongben Xu
Check 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 |
|---|---|---|---|
| Project Adam updates into the approximate null space of previous-task feature covariance | ResNet-18 continual learning · 10-split-CIFAR-100 | ACC (%): 73.77 vs 71.27 | Improved |
| Project Adam updates into the approximate null space of previous-task feature covariance | ResNet-18 continual learning · 20-split-CIFAR-100 | ACC (%): 75.95 vs 77.16 | Mixed |
| Project Adam updates into the approximate null space of previous-task feature covariance | ResNet-18 continual learning · 25-split-TinyImageNet | ACC (%): 58.28 vs 57.12 | Improved |
| Increase the approximate-null-space threshold a (ablation) | Adam-NSCL · 10-split-CIFAR-100, 20-split-CIFAR-100, and 25-split-TinyImageNet | BWT | Hurt |
| Increase the approximate-null-space threshold a (ablation) | Adam-NSCL · 20-split-CIFAR-100 and 25-split-TinyImageNet | ACC | Mixed |
| Use Adam-NSCL instead of OWM | ResNet-18 continual learning · 10-split-CIFAR-100 | ACC (%): +4.88% | Improved |
| Use Adam-NSCL instead of OWM | ResNet-18 continual learning · 20-split-CIFAR-100 | ACC (%): +7.48% | Improved |
| Use Adam-NSCL instead of OWM | ResNet-18 continual learning · 25-split-TinyImageNet | ACC (%): +8.3% | Improved |
Closest papers by meaning
- Training Networks in Null Space of Feature Covariance With Self-Supervision for Incremental Learning (TPAMI 2024 · similarity 0.92)
- Balancing Stability and Plasticity through Advanced Null Space in Continual Learning (ECCV 2022 · similarity 0.90)
- A continual learning survey: Defying forgetting in classification tasks (TPAMI 2021 · similarity 0.90)
- CRNet: A Fast Continual Learning Framework With Random Theory (TPAMI 2023 · similarity 0.90)
- Exploring Rationale Learning for Continual Graph Learning (AAAI 2025 · similarity 0.88)
- STAR: Stability-Inducing Weight Perturbation for Continual Learning (ICLR 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.
