Active One-shot Learning

TL;DR

This paper introduces a reinforcement learning approach to one-shot classification, enabling models to actively decide when to request labels, thereby improving efficiency and accuracy in streaming image classification tasks.

Contribution

It combines reinforcement learning with one-shot learning to enable active label requesting, a novel approach in the context of streaming image classification.

Findings

Model learns when to request labels effectively.

Achieves higher accuracy with fewer label requests.

Can trade accuracy for label efficiency.

Abstract

Recent advances in one-shot learning have produced models that can learn from a handful of labeled examples, for passive classification and regression tasks. This paper combines reinforcement learning with one-shot learning, allowing the model to decide, during classification, which examples are worth labeling. We introduce a classification task in which a stream of images are presented and, on each time step, a decision must be made to either predict a label or pay to receive the correct label. We present a recurrent neural network based action-value function, and demonstrate its ability to learn how and when to request labels. Through the choice of reward function, the model can achieve a higher prediction accuracy than a similar model on a purely supervised task, or trade prediction accuracy for fewer label requests.