Online Clustered Codebook

TL;DR

This paper introduces CVQ-VAE, an online clustering method for vector quantization that addresses codebook collapse by dynamically updating dead codevectors using encoded features as anchors, improving the capacity of VQ models.

Contribution

The paper proposes a simple, effective online clustering approach for VQ-VAE that mitigates codebook collapse and enhances codebook utilization in complex vision tasks.

Findings

Improved codebook utilization across various datasets.

Enhanced performance in reconstruction and generation tasks.

Easy integration into existing VQ models.

Abstract

Vector Quantisation (VQ) is experiencing a comeback in machine learning, where it is increasingly used in representation learning. However, optimizing the codevectors in existing VQ-VAE is not entirely trivial. A problem is codebook collapse, where only a small subset of codevectors receive gradients useful for their optimisation, whereas a majority of them simply ``dies off'' and is never updated or used. This limits the effectiveness of VQ for learning larger codebooks in complex computer vision tasks that require high-capacity representations. In this paper, we present a simple alternative method for online codebook learning, Clustering VQ-VAE (CVQ-VAE). Our approach selects encoded features as anchors to update the ``dead'' codevectors, while optimising the codebooks which are alive via the original loss. This strategy brings unused codevectors closer in distribution to the encoded features, increasing the likelihood of being chosen and optimized. We extensively validate the generalization capability of our quantiser on various datasets, tasks (e.g. reconstruction and generation), and architectures (e.g. VQ-VAE, VQGAN, LDM). Our CVQ-VAE can be easily integrated into the existing models with just a few lines of code.