VITS የድምፅ ፋይል
The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.
ርዕሱን በSSML መለያዎች ውስጥ ለጥሩ ቁጥጥር ይዞሩት:
<speak><prosody rate="slow">Slow speech</prosody></speak>
የተመረጠው ሞዴል የሚያውቃቸው መለያዎች - በጽሑፍዎ ውስጥ የሚከሰትበትን ቦታ ለመውሰድ ጠቅ ያድርጉ፦
ይህ ሞዴል ቀላል ጽሑፍን ያነባል፣ ስለዚህም በመስመር ውስጥ ያሉ ምልክቶች ይዘገያሉ፡፡ ለታክስ-ተኮር ስሜት እንደ ኦርፊየስ ወይም ባርክ ያሉ ግልጽ ሞዴሎችን ይለውጡ።
የራሱን ተናጋሪ ግለጽ (ቃል = ተናጋሪ):
ስለ 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.
ምርጥ ለ: General-purpose text-to-speech with natural prosody
ሁሉንም አጥፉ VITS ድምጾችበጥቂቱ
- የድር አዘጋጅ
- Jaehyeon Kim et al.
- ፈቃድ
- MIT
- ዐምድ
- free
- ፍጥነት
- fast
- የድምፅ ቅጂ
- አዎ
- ቋንቋዎች
- English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
- ፊደላት
- 2000