Melhubert: A Simplified Hubert On Mel Spectrograms

Abstract

Self-supervised models have had great success in learning speech representations that can generalize to various downstream tasks. However, most self-supervised models require a large amount of compute and multiple GPUs to train, significantly hampering the development of self-supervised learning. In an attempt to reduce the computation of training, we revisit the training of HuBERT, a highly successful self-supervised model. We improve and simplify several key components, including the loss function, input representation, and training in multiple stages. Our model, MelHuBERT, is able to achieve favorable performance on phone recognition, speaker identification, and automatic speech recognition against HuBERT, while saving 31.2% of the pre-training time, or equivalently 33.5% MACs per one second speech. The code and pre-trained models are available in https://github.com/nervjack2/MelHuBERT.

Stats

• citations0
• S2 citations—
• github stars70
• HF likes0
• heat score3.70
• arxiv keylin2022melhubert

Code

• nervjack2/MelHuBERT—

Related papers

• Fast-hubert: An Efficient Training Framework For Self-supervised Speech Representation Learning (2023)0.00
• Ms-hubert: Mitigating Pre-training And Inference Mismatch In Masked Language Modelling Methods For Learning Speech Representations (2024)4.52
• Lighthubert: Lightweight And Configurable Speech Representation Learning With Once-for-all Hidden-unit BERT (2022)15.51
• Hubert: Self-supervised Speech Representation Learning By Masked Prediction Of Hidden Units (2021)25.30
• Distilhubert: Speech Representation Learning By Layer-wise Distillation Of Hidden-unit BERT (2021)15.06
• Fithubert: Going Thinner And Deeper For Knowledge Distillation Of Speech Self-supervised Learning (2022)10.97
• Selective Hubert: Self-supervised Pre-training For Target Speaker In Clean And Mixture Speech (2023)7.81
• Cocktail Hubert: Generalized Self-supervised Pre-training For Mixture And Single-source Speech (2023)6.77