English

Knowledge Distillation with the Reused Teacher Classifier

Computer Vision and Pattern Recognition 2022-03-29 v1

Abstract

Knowledge distillation aims to compress a powerful yet cumbersome teacher model into a lightweight student model without much sacrifice of performance. For this purpose, various approaches have been proposed over the past few years, generally with elaborately designed knowledge representations, which in turn increase the difficulty of model development and interpretation. In contrast, we empirically show that a simple knowledge distillation technique is enough to significantly narrow down the teacher-student performance gap. We directly reuse the discriminative classifier from the pre-trained teacher model for student inference and train a student encoder through feature alignment with a single 2\ell_2 loss. In this way, the student model is able to achieve exactly the same performance as the teacher model provided that their extracted features are perfectly aligned. An additional projector is developed to help the student encoder match with the teacher classifier, which renders our technique applicable to various teacher and student architectures. Extensive experiments demonstrate that our technique achieves state-of-the-art results at the modest cost of compression ratio due to the added projector.

Keywords

Cite

@article{arxiv.2203.14001,
  title  = {Knowledge Distillation with the Reused Teacher Classifier},
  author = {Defang Chen and Jian-Ping Mei and Hailin Zhang and Can Wang and Yan Feng and Chun Chen},
  journal= {arXiv preprint arXiv:2203.14001},
  year   = {2022}
}

Comments

Accepted to CVPR-2022

R2 v1 2026-06-24T10:26:41.983Z