Abstract:Deep neural models have achieved remarkable performance on various supervised and unsupervised learning tasks, but it is a challenge to deploy these large-size networks on resource-limited devices. As a representative type of model compression and acceleration methods, knowledge distillation (KD) solves this problem by transferring knowledge from heavy teachers to lightweight students. However, most distillation methods focus on imitating the responses of teacher networks but ignore the information redundancy of student networks. In this article, we propose a novel distillation framework difference-based channel contrastive distillation (DCCD), which introduces channel contrastive knowledge and dynamic difference knowledge into student networks for redundancy reduction. At the feature level, we construct an efficient contrastive objective that broadens student networks' feature expression space and preserves richer information in the feature extraction stage. At the final output level, more detailed knowledge is extracted from teacher networks by making a difference between multiview augmented responses of the same instance. We enhance student networks to be more sensitive to minor dynamic changes. With the improvement of two aspects of DCCD, the student network gains contrastive and difference knowledge and reduces its overfitting and redundancy. Finally, we achieve surprising results that the student approaches and even outperforms the teacher in test accuracy on CIFAR-100. We reduce the top-1 error to 28.16% on ImageNet classification and 24.15% for cross-model transfer with ResNet-18. Empirical experiments and ablation studies on popular datasets show that our proposed method can achieve state-of-the-art accuracy compared with other distillation methods.

Knowledge Distillation for Machine Translation

DCCD: Reducing Neural Network Redundancy Via Distillation

Dual Knowledge Distillation for neural machine translation

Multilingual Neural Machine Translation with Knowledge Distillation

Continual Knowledge Distillation for Neural Machine Translation

Life-long Learning for Multilingual Neural Machine Translation with Knowledge Distillation

Exploring All-In-One Knowledge Distillation Framework for Neural Machine Translation

Nearest Neighbor Knowledge Distillation for Neural Machine Translation.

Unraveling Key Factors of Knowledge Distillation

Selective Knowledge Distillation for Non-Autoregressive Neural Machine Translation

Towards Understanding and Improving Knowledge Distillation for Neural Machine Translation

Continual Learning with Confidence-based Multi-teacher Knowledge Distillation for Neural Machine Translation

A Model Compression Method Using Significant Data and Knowledge Distillation

Building a Multi-domain Neural Machine Translation Model using Knowledge Distillation

Joint architecture and knowledge distillation in CNN for Chinese text recognition

Weight Distillation: Transferring the Knowledge in Neural Network Parameters

Self-Knowledge Distillation in Natural Language Processing

Selective Knowledge Distillation for Neural Machine Translation

Knowledge Distillation Application Technology for Chinese NLP

Why Skip If You Can Combine: A Simple Knowledge Distillation Technique for Intermediate Layers

Knowledge Distillation for Multilingual Unsupervised Neural Machine Translation