On the Soft-Subnetwork for Few-shot Class Incremental Learning
About
Inspired by Regularized Lottery Ticket Hypothesis (RLTH), which hypothesizes that there exist smooth (non-binary) subnetworks within a dense network that achieve the competitive performance of the dense network, we propose a few-shot class incremental learning (FSCIL) method referred to as \emph{Soft-SubNetworks (SoftNet)}. Our objective is to learn a sequence of sessions incrementally, where each session only includes a few training instances per class while preserving the knowledge of the previously learned ones. SoftNet jointly learns the model weights and adaptive non-binary soft masks at a base training session in which each mask consists of the major and minor subnetwork; the former aims to minimize catastrophic forgetting during training, and the latter aims to avoid overfitting to a few samples in each new training session. We provide comprehensive empirical validations demonstrating that our SoftNet effectively tackles the few-shot incremental learning problem by surpassing the performance of state-of-the-art baselines over benchmark datasets.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Few-Shot Class-Incremental Learning | miniImageNet (test) | Accuracy (Session 1)75.08 | 173 | |
| Few-Shot Class-Incremental Learning | CIFAR100 (test) | Session 4 Top-1 Acc66.14 | 122 | |
| Few-Shot Class-Incremental Learning | CUB200 (test) | Accuracy (Session 1)74.58 | 92 | |
| Few-Shot Class-Incremental Learning | CUB-200 | Session 1 Accuracy74.58 | 75 | |
| Few-Shot Class-Incremental Learning | CIFAR100 | Accuracy (S0)79.88 | 67 | |
| Few-Shot Class-Incremental Learning | CUB200 (incremental sessions) | Session 0 Accuracy78.07 | 37 | |
| Few-Shot Class-Incremental Learning | miniImageNet 5-way 5-shot (incremental) | Accuracy S077.17 | 21 | |
| Few-Shot Class-Incremental Learning | CUB-200 | Session 0 Accuracy78.14 | 21 | |
| Few-Shot Class-Incremental Learning | CIFAR-100 60 base classes, 5-way 5-shot, Session 0 | Accuracy79.88 | 20 | |
| Few-Shot Class-Incremental Learning | CIFAR-100 60 base classes 5-way 5-shot Session 1 | Accuracy75.54 | 20 |