Distilling Holistic Knowledge with Graph Neural Networks
This work addresses the challenge of improving knowledge distillation for training smaller neural networks, presenting an incremental advancement by combining existing knowledge types in a novel way.
The paper tackles the problem of knowledge distillation by proposing a method to integrate individual and relational knowledge while preserving their inherent correlations, using graph neural networks to represent holistic knowledge and distilling it contrastively, achieving effectiveness demonstrated through experiments on benchmark datasets.
Knowledge Distillation (KD) aims at transferring knowledge from a larger well-optimized teacher network to a smaller learnable student network.Existing KD methods have mainly considered two types of knowledge, namely the individual knowledge and the relational knowledge. However, these two types of knowledge are usually modeled independently while the inherent correlations between them are largely ignored. It is critical for sufficient student network learning to integrate both individual knowledge and relational knowledge while reserving their inherent correlation. In this paper, we propose to distill the novel holistic knowledge based on an attributed graph constructed among instances. The holistic knowledge is represented as a unified graph-based embedding by aggregating individual knowledge from relational neighborhood samples with graph neural networks, the student network is learned by distilling the holistic knowledge in a contrastive manner. Extensive experiments and ablation studies are conducted on benchmark datasets, the results demonstrate the effectiveness of the proposed method. The code has been published in https://github.com/wyc-ruiker/HKD