Audio self-supervised learning: A survey