VITS TTS
The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.
Avvolgi il tuo testo nei tag SSML per un controllo preciso:
<speak><prosody rate="slow">Slow speech</prosody></speak>
Tags il modello selezionato comprende clic su
Questo modello legge testo semplice, quindi i tag inline vengono ignorati. Per le emozioni basate sui tag, passare a un modello espressivo come Orpheus o Bark.
Definire le pronunciazioni personalizzate (parola = pronuncia):
Informazioni 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.
Meglio per: General-purpose text-to-speech with natural prosody
Sfoglia tutti VITS vociA colpo d'occhio
- Sviluppatore
- Jaehyeon Kim et al.
- Licenza
- MIT
- Livello
- free
- Velocità
- fast
- Clonazione vocale
- No.
- Lingue
- English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
- Caratteri massimi
- 2000