Study
Generative AI · Text-to-speech
Text-to-speech with a small end-to-end model (Meta's MMS-TTS English, a VITS model): text to waveform in one network, plus a look at the generated audio's waveform and spectrogram.
- VITS-style models go from text straight to a waveform in one network, with no separate vocoder step.
- A spectrogram shows speech as energy over time and frequency; it's the representation most speech models work with.