Student Helping Teacher: Teacher Evolution via Self-Knowledge Distillation

TL;DR

This paper introduces TESKD, a novel approach where multiple hierarchical students help improve the teacher network through shared features, leading to better performance on standard benchmarks.

Contribution

The paper proposes a new teacher evolution method via self-knowledge distillation, enabling the teacher to learn from multiple students sharing the same backbone.

Findings

ResNet-18 achieves 79.15% accuracy on CIFAR-100, outperforming baseline by 4.74%.

ResNet-18 achieves 71.14% accuracy on ImageNet, outperforming baseline by 1.43%.

The method demonstrates significant improvements across various network settings.

Abstract

Knowledge distillation usually transfers the knowledge from a pre-trained cumbersome teacher network to a compact student network, which follows the classical teacher-teaching-student paradigm. Based on this paradigm, previous methods mostly focus on how to efficiently train a better student network for deployment. Different from the existing practices, in this paper, we propose a novel student-helping-teacher formula, Teacher Evolution via Self-Knowledge Distillation (TESKD), where the target teacher (for deployment) is learned with the help of multiple hierarchical students by sharing the structural backbone. The diverse feedback from multiple students allows the teacher to improve itself through the shared feature representations. The effectiveness of our proposed framework is demonstrated by extensive experiments with various network settings on two standard benchmarks including CIFAR-100 and ImageNet. Notably, when trained together with our proposed method, ResNet-18 achieves 79.15% and 71.14% accuracy on CIFAR-100 and ImageNet, outperforming the baseline results by 4.74% and 1.43%, respectively. The code is available at: https://github.com/zhengli427/TESKD.