VITS TTS
The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.
Cuir do théacs i gclibeanna SSML le haghaidh rialú beacht:
<speak><prosody rate="slow">Slow speech</prosody></speak>
Clibeanna a thuigeann an tsamhail roghnaithe — cliceáil chun ceann a scaoileadh isteach i do théacs nuair a tharlaíonn sé:
Léann an tsamhail seo gnáth- théacs, mar sin déantar neamhaird ar chlibeanna inlíne. Chun mothúchán clibbhunaithe a chruthú, athraigh go samhail léiritheach cosúil le Orpheus nó Bark.
Sainmhínigh fuaimniú saincheaptha (focal = fuaimniú):
Eolas Faoi 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.
Is Fearr le haghaidh: General-purpose text-to-speech with natural prosody
Brabhsáil Uile VITS guthannaAg Sracfhéachaint
- Forbróir
- Jaehyeon Kim et al.
- Ceadúnas
- MIT
- Tír
- free
- Luas
- fast
- Clónáil gutha
- & Ná Sábháil
- Teangacha
- English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
- Carachtair Uasta
- 2000