Channel Importance Matters in Few-Shot Image Classification

TL;DR

This paper demonstrates that a simple channel-wise feature transformation significantly enhances the generalization of vision models in Few-Shot Learning by addressing channel bias issues, revealing a core challenge in model transferability.

Contribution

It introduces a channel-wise feature transformation method that improves FSL generalization and analyzes the impact of channel bias on representation transfer.

Findings

Channel bias causes transfer difficulties in FSL.

Channel-wise transformation improves test-time generalization.

CNNs respond incorrectly to channel importance shifts.

Abstract

Few-Shot Learning (FSL) requires vision models to quickly adapt to brand-new classification tasks with a shift in task distribution. Understanding the difficulties posed by this task distribution shift is central to FSL. In this paper, we show that a simple channel-wise feature transformation may be the key to unraveling this secret from a channel perspective. When facing novel few-shot tasks in the test-time datasets, this transformation can greatly improve the generalization ability of learned image representations, while being agnostic to the choice of training algorithms and datasets. Through an in-depth analysis of this transformation, we find that the difficulty of representation transfer in FSL stems from the severe channel bias problem of image representations: channels may have different importance in different tasks, while convolutional neural networks are likely to be insensitive, or respond incorrectly to such a shift. This points out a core problem of the generalization ability of modern vision systems and needs further attention in the future. Our code is available at https://github.com/Frankluox/Channel_Importance_FSL.