Isikhulumi esihlukanisiwe esingenalutho

Uma abantu ababi bakhuluma ngabanye ku-track eyodwa, le thuluzi lisusa umsindo onamandla phambili futhi linciphisa ukungqubuzana, kwenza ukuxoxa okuphakathi kube lula ukulandelela nokudlulisa.

Imodeli ye-AI Lezi zithuluzi ziqhuba imodeli ye-AI ekhiqizayo kwi-GPU, ngakho-ke zithatha isikhathi eside kunalezi zihlungi ezisheshayo. Zikhululekile ukuzama; ama-plans akhokhelwayo athola isiqalo ku-GPU.

🎧

Thola ihele lomsindo noma ihele levidiyo lapha

noma

MP3, WAV, M4A, FLAC, OGG, AAC

Ukuhlanza umsindo wakho…

Imiphumela yakho ixhunywe
Phambi

Usizo: Chofoza isikhala sokuvula/savala phambi/emuva.

Layisha phezulu izimpande
Layisha ngezansi umsindo ohlanziwe Bhala kuhele obukhulu

Indlela esebenza ngayo

Ukukhuluma okuguquguqukayo kuhlukaniswa ngomthombo ukuze umsindo ongaphambili uhlukaniswe noเสียงo negumbi. Imiphumela izoba umsindo ocacile okhuluma kuphela kusuka ku-cross-talk recording.

Isebenza kanjani

  • Ukuxhumana okuqondile emibuzweni
  • Ukurekhodwa kwe-microphone eyodwa-ne-abantu abathathu
  • Ukuhlanza umsindo wezingxoxo
  • Ukuhlela ukudluliswa kokubhala

Iminingwane

Injini
Mel-Band RoFormer
Amafomethi
MP3, WAV, M4A, FLAC, OGG, AAC
Intengo
Ikhululekile ukuzama

Imibuzo ebuzwa kaningi

Ihlukanisa umsindo ohamba phambili ohamba phambili ohamba phambili futhi icindezela ukuguquguquka. Uhlu oluphelele lwe-speaker-per-diaarization kuma-tracks ahlukene lukhona ku-roadmap yethu; namuhla ihlanza umsindo ongaphambili.

Yebo - ukuncipha kwezwi elincintisana negumbi kwenza ukukhuluma-nokubhala kube lula kakhulu kungxenyekazi ezixhumene.

Ukurekhoda lapho umsindo wakho ofuna ukuwuthola uphakathi kwe-mic ehlukile kahle, njengoba ukufinyelela kunikeza imodeli i-foreground cue enamandla.

Ukuhlukaniswa kwezwi kufakwe izizwi ezimbili ezikhuluma ngazo ngazo zonke izilimi ezikhona, ngenkathi ukukhishwa kwezwi eligunyaziwe kususa isizwi esisodwa kusuka ku-mix ebanzi yamazwi, umculo nezingcingo.

Ilungele izizwi ezimbili ezixhumene; nge-3 noma ngaphezulu iphinde ikhuphule umsindo ophakathi nendawo, kodwa imiphumela ayihlanzekile kune-cross-talk recording esemthethweni yomntu omibili.

Isiqephu esijwayelekile sokuxhumana siqhubekela phambili ngaphansi kwemingwane, silinganisela ngesikhathi sokuzisakaza ngaphezu kokuthi kuningi kangakanani ukungqubuzana.

Izizwi ezilinganayo-volume ziyinkinga enzima futhi zishiya okuningi kokhulumayo odlalayo; ithuluzi lisebenza kahle uma okudingayo kucacile, isizwi esiphakeme.

Amathuluzi ahlobene