Prototypical Contrastive Learning of Unsupervised Representations

TL;DR

Prototypical Contrastive Learning (PCL) introduces prototypes into unsupervised representation learning, improving semantic encoding and outperforming existing methods on benchmarks, especially in low-resource transfer scenarios.

Contribution

The paper proposes PCL, a novel unsupervised learning method that integrates prototypes with contrastive learning using an EM framework, enhancing semantic structure encoding.

Findings

PCL outperforms state-of-the-art contrastive methods on multiple benchmarks.

PCL shows substantial improvements in low-resource transfer learning.

The ProtoNCE loss generalizes InfoNCE for prototype-based contrastive learning.

Abstract

This paper presents Prototypical Contrastive Learning (PCL), an unsupervised representation learning method that addresses the fundamental limitations of instance-wise contrastive learning. PCL not only learns low-level features for the task of instance discrimination, but more importantly, it implicitly encodes semantic structures of the data into the learned embedding space. Specifically, we introduce prototypes as latent variables to help find the maximum-likelihood estimation of the network parameters in an Expectation-Maximization framework. We iteratively perform E-step as finding the distribution of prototypes via clustering and M-step as optimizing the network via contrastive learning. We propose ProtoNCE loss, a generalized version of the InfoNCE loss for contrastive learning, which encourages representations to be closer to their assigned prototypes. PCL outperforms state-of-the-art instance-wise contrastive learning methods on multiple benchmarks with substantial improvement in low-resource transfer learning. Code and pretrained models are available at https://github.com/salesforce/PCL.