Multi-granularity for knowledge distillation

Abstract

Considering the fact that students have different abilities to understand the knowledge imparted by teachers, a multi-granularity distillation mechanism is proposed for transferring more understandable knowledge for student networks. A multi-granularity self-analyzing module of the teacher network is designed, which enables the student network to learn knowledge from different teaching patterns. Furthermore, a stable excitation scheme is proposed for robust supervision for the student training. The proposed distillation mechanism can be embedded into different distillation frameworks, which are taken as baselines. Experiments show the mechanism improves the accuracy by 0.58% on average and by 1.08% in the best over the baselines, which makes its performance superior to the state-of-the-arts. It is also exploited that the student's ability of fine-tuning and robustness to noisy inputs can be improved via the proposed mechanism. The code is available at https://github.com/shaoeric/multi-granularity-distillation.

14 pages, 12 figures

References in corpus (8)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Similarity of Neural Network Representations Revisited
• Like What You Like: Knowledge Distill via Neuron Selectivity Transfer
• Torchreid: A Library for Deep Learning Person Re-Identification in Pytorch
• Rethinking Soft Labels for Knowledge Distillation: A Bias-Variance Tradeoff Perspective
• Knowledge Distillation Methods for Efficient Unsupervised Adaptation Across Multiple Domains
• PyTorch Metric Learning
• Show, Attend and Distill:Knowledge Distillation via Attention-based Feature Matching