English

Knowledge Distillation via Token-level Relationship Graph

Machine Learning 2023-06-23 v1 Artificial Intelligence

Abstract

Knowledge distillation is a powerful technique for transferring knowledge from a pre-trained teacher model to a student model. However, the true potential of knowledge transfer has not been fully explored. Existing approaches primarily focus on distilling individual information or instance-level relationships, overlooking the valuable information embedded in token-level relationships, which may be particularly affected by the long-tail effects. To address the above limitations, we propose a novel method called Knowledge Distillation with Token-level Relationship Graph (TRG) that leverages the token-wise relational knowledge to enhance the performance of knowledge distillation. By employing TRG, the student model can effectively emulate higher-level semantic information from the teacher model, resulting in improved distillation results. To further enhance the learning process, we introduce a token-wise contextual loss called contextual loss, which encourages the student model to capture the inner-instance semantic contextual of the teacher model. We conduct experiments to evaluate the effectiveness of the proposed method against several state-of-the-art approaches. Empirical results demonstrate the superiority of TRG across various visual classification tasks, including those involving imbalanced data. Our method consistently outperforms the existing baselines, establishing a new state-of-the-art performance in the field of knowledge distillation.

Keywords

Cite

@article{arxiv.2306.12442,
  title  = {Knowledge Distillation via Token-level Relationship Graph},
  author = {Shuoxi Zhang and Hanpeng Liu and Kun He},
  journal= {arXiv preprint arXiv:2306.12442},
  year   = {2023}
}
R2 v1 2026-06-28T11:11:02.059Z