SENet: A Spectral Filtering Approach to Represent Exemplars for Few-shot Learning
This work addresses the issue of underfitting in prototype-based few-shot learning for computer vision tasks, offering a novel method to better represent categories using exemplars.
The paper tackles the problem of prototype representation in few-shot learning by proposing SENet, which uses spectral filtering to shrink sample embeddings towards their mean, resulting in improved classification performance on miniImageNet, tiered-ImageNet, and CIFAR-FS datasets.
Prototype is widely used to represent internal structure of category for few-shot learning, which was proposed as a simple inductive bias to address the issue of overfitting. However, since prototype representation is normally averaged from individual samples, it can appropriately to represent some classes but with underfitting to represent some others that can be batter represented by exemplars. To address this problem, in this work, we propose Shrinkage Exemplar Networks (SENet) for few-shot classification. In SENet, categories are represented by the embedding of samples that shrink towards their mean via spectral filtering. Furthermore, a shrinkage exemplar loss is proposed to replace the widely used cross entropy loss for capturing the information of individual shrinkage samples. Several experiments were conducted on miniImageNet, tiered-ImageNet and CIFAR-FS datasets. The experimental results demonstrate the effectiveness of our proposed method.