Prototypical Contrastive Learning of Unsupervised Representations
This addresses the problem of learning semantic representations without labels for machine learning practitioners, offering a novel method that is not incremental.
The paper tackles the limitations of instance-wise contrastive learning in unsupervised representation learning by introducing Prototypical Contrastive Learning (PCL), which implicitly encodes semantic structures into embeddings and achieves substantial improvements in low-resource transfer learning on multiple benchmarks.
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.