Knowledge Distillation via Token-level Relationship Graph
Zhang, Shuoxi, Liu, Hanpeng, He, Kun
–arXiv.org Artificial Intelligence
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.
arXiv.org Artificial Intelligence
Jun-20-2023
- Country:
- North America > United States
- New York (0.04)
- Asia > China
- Hubei Province > Wuhan (0.04)
- North America > United States
- Genre:
- Research Report > New Finding (1.00)
- Industry:
- Education (1.00)
- Technology: