Thoughts on audio AI, speech processing, and research.
Two routes to audio generation — discrete codec + LM vs. continuous diffusion / flow — how they converge, and the theory underneath.
My first blog post — why I'm starting to write about audio AI and speech processing.