Margin-Based Few-Shot Class-Incremental Learning with Class-Level Overfitting Mitigation
About
Few-shot class-incremental learning (FSCIL) is designed to incrementally recognize novel classes with only few training samples after the (pre-)training on base classes with sufficient samples, which focuses on both base-class performance and novel-class generalization. A well known modification to the base-class training is to apply a margin to the base-class classification. However, a dilemma exists that we can hardly achieve both good base-class performance and novel-class generalization simultaneously by applying the margin during the base-class training, which is still under explored. In this paper, we study the cause of such dilemma for FSCIL. We first interpret this dilemma as a class-level overfitting (CO) problem from the aspect of pattern learning, and then find its cause lies in the easily-satisfied constraint of learning margin-based patterns. Based on the analysis, we propose a novel margin-based FSCIL method to mitigate the CO problem by providing the pattern learning process with extra constraint from the margin-based patterns themselves. Extensive experiments on CIFAR100, Caltech-USCD Birds-200-2011 (CUB200), and miniImageNet demonstrate that the proposed method effectively mitigates the CO problem and achieves state-of-the-art performance.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Few-Shot Class-Incremental Learning | miniImageNet (test) | Accuracy (Session 1)68.09 | 173 | |
| Few-Shot Class-Incremental Learning | CIFAR100 (test) | Session 4 Top-1 Acc59.26 | 122 | |
| Few-Shot Class-Incremental Learning | CIFAR100 | Accuracy (S0)74.2 | 67 | |
| Few-Shot Class-Incremental Learning | miniImageNet 5-way 5-shot (incremental) | Accuracy S073.08 | 21 | |
| Few-Shot Class-Incremental Learning | miniImageNet 60 base classes 5-way 5-shot (incremental) | Session 0 Accuracy73.08 | 20 | |
| Few-Shot Class-Incremental Learning | CIFAR-100 60 base classes, 5-way 5-shot, Session 0 | Accuracy74.2 | 20 | |
| Few-Shot Class-Incremental Learning | CIFAR-100 60 base classes 5-way 5-shot Session 8 | Accuracy50.25 | 20 | |
| Few-Shot Class-Incremental Learning | CIFAR-100 60 base classes 5-way 5-shot Session 1 | Accuracy69.83 | 20 | |
| Few-Shot Class-Incremental Learning | CIFAR-100 60 base classes 5-way 5-shot Session 2 | Accuracy66.17 | 20 | |
| Few-Shot Class-Incremental Learning | CIFAR-100 60 base classes 5-way 5-shot Session 3 | Accuracy62.39 | 20 |