VITS TTS
The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.
Aplauzt savu tekstu SSML tagus precīzai kontrolei:
<speak><prosody rate="slow">Slow speech</prosody></speak>
Tags izvēlētais modelis saprot — noklikšķiniet, lai iemestu vienu jūsu tekstā, kur tas notiek:
Šis modelis lasa vienkāršu tekstu, tāpēc tiek ignorēti inline tagi. Uz tag-based emocijas, pāriet uz izteiksmīgu modeli, piemēram, Orpheus vai Bark.
Definēt pielāgotu izrunas (vārds = izruna):
Par VITS
VITS — Variational Inference with adversarial learning for end-to-end Text-to-Speech — was introduced by Jaehyeon Kim and collaborators in 2021 and became a foundational architecture for modern neural speech. Rather than the older two-stage pipeline, it synthesizes audio in a single parallel end-to-end pass, pairing a variational autoencoder with normalizing flows and a GAN-style adversarial training process to lift naturalness. At about 25M parameters and trained on ~585 hours, it produces natural prosody at fast inference speeds and supports multiple speakers. It serves as a solid general-purpose, free baseline and underpins many later models such as Piper and MeloTTS.
Labākais: General-purpose text-to-speech with natural prosody
Pārlūkot visu VITS balsisĪsumā
- Izstrādātājs
- Jaehyeon Kim et al.
- Licence
- MIT
- Līmeņrādis
- free
- Ātrums
- fast
- Balss klonēšana
- Nē
- Valodas
- English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
- Maks. rakstzīmes
- 2000