Open Implementation and Study of BEST-RQ for Speech Processing

TL;DR

This paper re-implements and evaluates BEST-RQ, a simpler SSL method for speech processing, demonstrating comparable performance to wav2vec 2.0 while significantly reducing training time across multiple tasks.

Contribution

It provides an open-source implementation of BEST-RQ and compares its performance and efficiency to wav2vec 2.0 on various speech tasks.

Findings

BEST-RQ achieves similar downstream task performance as wav2vec 2.0.

Training time for BEST-RQ is over twice as fast as wav2vec 2.0.

The study offers detailed implementation insights and a preliminary evaluation across four tasks.

Abstract

Self-Supervised Learning (SSL) has proven to be useful in various speech tasks. However, these methods are generally very demanding in terms of data, memory, and computational resources. BERT-based Speech pre-Training with Random-projection Quantizer (BEST-RQ), is an SSL method that has shown great performance on Automatic Speech Recognition (ASR) while being simpler than other SSL methods, such as wav2vec 2.0. Despite BEST-RQ's great performance, details are lacking in the original paper, such as the amount of GPU/TPU hours used in pre-training, and there is no official easy-to-use open-source implementation. Furthermore, BEST-RQ has not been evaluated on other downstream tasks aside from ASR and speech translation. In this work, we describe a re-implementation of a Random-projection quantizer and perform a preliminary study with a comparison to wav2vec 2.0 on four downstream tasks. We discuss the details and differences of our implementation. We show that a random projection quantizer can achieve similar downstream performance as wav2vec 2.0 while decreasing training time by over a factor of two.