Analyzing, transcribing, generating, or understanding audio and speech signals.
No models in this category are strong on that dimension yet.