Introspective Learning by Distilling Knowledge from Online Self-explanation

TL;DR

This paper introduces a novel introspective learning method that distills knowledge from a model's own online self-explanations, enhancing learning without external teachers or peer networks.

Contribution

It proposes a new approach to improve neural network training by leveraging self-explanations as privileged information, inspired by knowledge distillation techniques.

Findings

Models trained with introspective learning outperform standard training.

The method achieves competitive results without external teachers.

Incorporating class-similarity information enhances knowledge transfer.

Abstract

In recent years, many explanation methods have been proposed to explain individual classifications of deep neural networks. However, how to leverage the created explanations to improve the learning process has been less explored. As the privileged information, the explanations of a model can be used to guide the learning process of the model itself. In the community, another intensively investigated privileged information used to guide the training of a model is the knowledge from a powerful teacher model. The goal of this work is to leverage the self-explanation to improve the learning process by borrowing ideas from knowledge distillation. We start by investigating the effective components of the knowledge transferred from the teacher network to the student network. Our investigation reveals that both the responses in non-ground-truth classes and class-similarity information in teacher's outputs contribute to the success of the knowledge distillation. Motivated by the conclusion, we propose an implementation of introspective learning by distilling knowledge from online self-explanations. The models trained with the introspective learning procedure outperform the ones trained with the standard learning procedure, as well as the ones trained with different regularization methods. When compared to the models learned from peer networks or teacher networks, our models also show competitive performance and requires neither peers nor teachers.