Wasserstein Contrastive Representation Distillation

Liqun Chen, Zhe Gan, Dong Wang, Jingjing Liu, Ricardo Henao, Lawrence Carin

Research output: Contribution to journalArticlepeer-review

Abstract

The primary goal of knowledge distillation (KD) is to encapsulate the information of a model learned from a teacher network into a student network, with the latter being more compact than the former. Existing work, e.g., using Kullback-Leibler divergence for distillation, may fail to capture important structural knowledge in the teacher network and often lacks the ability for feature generalization, particularly in situations when teacher and student are built to address different classification tasks. We propose Wasserstein Contrastive Representation Distillation (WCoRD), which leverages both primal and dual forms of Wasserstein distance for KD. The dual form is used for global knowledge transfer, yielding a contrastive learning objective that maximizes the lower bound of mutual information between the teacher and the student networks. The primal form is used for local contrastive knowledge transfer within a mini-batch, effectively matching the distributions of features between the teacher and the student networks. Experiments demonstrate that the proposed WCoRD method outperforms state-of-the-art approaches on privileged information distillation, model compression and cross-modal transfer.
Original languageEnglish (US)
JournalArxiv preprint
StatePublished - Dec 15 2020
Externally publishedYes

Keywords

  • cs.LG
  • cs.CV

Fingerprint Dive into the research topics of 'Wasserstein Contrastive Representation Distillation'. Together they form a unique fingerprint.

Cite this