Materials
Long-form notes and preparation documents.
-
Speech-to-speech modelling
How modern speech-to-speech and audio-visual generation systems are built in 2026: speech input and ASR, codecs, autoregressive and flow-matching generation, full-duplex models, steering, post-training, audio-visual diffusion, multimodal LLMs, and system design.
-
Speech-to-speech: quick read
Condensed one-page version of Speech-to-speech modelling: the key ideas, models and interview lines per chapter, readable in about 20 minutes.