VITS ТТС
The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.
Умотајте текст у ССМЛ ознаке за прецизну контролу:
<speak><prosody rate="slow">Slow speech</prosody></speak>
Означе изабрани модел разуме — кликните да га испустите у текст где се то дешава:
Овај модел чита обичан текст, па се ознаке за успостављене на линији игноришу. За емоције засноване на ознакама пребаците на изражавајући модел попут Орфеја или Барка.
Дефинишите посебне изговоре (слов = изговор):
О 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.
Најбоље за: General-purpose text-to-speech with natural prosody
Прегледај све VITS гласовиНа један поглед
- Програмер
- Jaehyeon Kim et al.
- Лиценца
- MIT
- Низ
- free
- Брзина
- fast
- Гласово клонирање
- Не.
- језици
- English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
- Макс. знакова
- 2000