# Self-supervised speech learning

HuBERT is a self-supervised speech representation learning model that learns speech features through BERT-like prediction loss, suitable for tasks such as speech recognition.