[2406.08800] Can Synthetic Audio From Generative Foundation Models Assist Audio Recognition and Speech Modeling?