Few-shot Learning by Exploiting Visual Concepts within CNNs

TL;DR

This paper introduces interpretable models for few-shot learning that leverage visual concepts within CNNs, enabling effective recognition from limited data while enhancing interpretability.

Contribution

The authors develop novel models based on visual concepts in CNNs, demonstrating improved interpretability and competitive performance in few-shot learning tasks.

Findings

Models achieve competitive accuracy in few-shot learning.

Visual concepts reveal CNNs' natural capability for few-shot recognition.

Proposed models are more flexible and interpretable than existing methods.

Abstract

Convolutional neural networks (CNNs) are one of the driving forces for the advancement of computer vision. Despite their promising performances on many tasks, CNNs still face major obstacles on the road to achieving ideal machine intelligence. One is that CNNs are complex and hard to interpret. Another is that standard CNNs require large amounts of annotated data, which is sometimes hard to obtain, and it is desirable to learn to recognize objects from few examples. In this work, we address these limitations of CNNs by developing novel, flexible, and interpretable models for few-shot learning. Our models are based on the idea of encoding objects in terms of visual concepts (VCs), which are interpretable visual cues represented by the feature vectors within CNNs. We first adapt the learning of VCs to the few-shot setting, and then uncover two key properties of feature encoding using VCs, which we call category sensitivity and spatial pattern. Motivated by these properties, we present two intuitive models for the problem of few-shot learning. Experiments show that our models achieve competitive performances, while being more flexible and interpretable than alternative state-of-the-art few-shot learning methods. We conclude that using VCs helps expose the natural capability of CNNs for few-shot learning.