Diarization Speaker
Diarization na lasifikar yana amsa tambayar "wane yayi magana yaushe?" ta hanyar rarraba rikodin sauti zuwa sassan da aka yiwa lakabi da asalin lasifikar.
Dubawa
It turns a single stream of mixed voices into a timeline showing exactly which person was talking at each moment.
Zurfafa nutsewa
Diarization yana aiwatar da sauti a matakai. Na farko, gano ayyukan murya yana gano yankunan magana. Daga nan sai a yanke jawabin zuwa gajerun sassa, kuma kowane bangare yana jujjuya shi zuwa wani tsayayyen tsayayyen vector da ake kira mai magana da magana (a tarihi i-vectors ko x-vectors, a yanzu yawanci abubuwan sakawa neural kamar ECAPA-TDNN). Matakan tari (gungu-gungu ko gungu na gani) ƙungiyoyi masu nau'in haɗawa iri ɗaya cikin masu magana, galibi ba tare da sanin adadin masu magana a gaba ba. A ƙarshe, ana tace iyakoki kuma ana warware maganganun da suka mamaye juna. Mahimmanci, diarization ba ya buƙatar sanin ko su wane ne mutanen; kawai yana sanya alamun da ba a san su ba kamar "Speaker 1" da "Speaker 2." Ana auna daidaito tare da Kuskuren Diarization (DER), wanda ya haɗa maganganun da aka rasa, ƙararrawar ƙarya, da ruɗin magana.
Fahimtar Fasaha
Babban dabarar ita ce haɗa lasifika: hanyar sadarwa ta jijiyoyi da aka horar da su ta yadda faifan bidiyo daga mutum ɗaya suka sauka kusa da juna a sararin samaniya da faifan bidiyo daga mutane daban-daban suna ƙasa nesa. Tari sannan yana aiki akan waɗannan abubuwan haɗin gwiwa maimakon ɗanyen sauti. Zamani "ƙarshe-zuwa-ƙarshen jijiya diarization" (EEND) ya maye gurbin tari tare da hanyar sadarwa guda ɗaya ta amfani da horarwar da ba ta dace ba, wacce ke sarrafa maganganun da suka wuce gona da iri fiye da tarin bututun mai kawai waɗanda ke ɗaukar mai magana ɗaya lokaci guda.
Dabarun Tasiri
Shiga ku isa
Yana inganta samun dama ta hanyar rubutu, ba da labari, da mu'amalar murya.
Kudin da kasafin kuɗi
Ƙungiyoyin kafofin watsa labaru na iya jigilar sauti mai gogewa cikin sauri tare da ƙaramin kasafin kuɗi.
Gudu da sikelin
Tsarin fuskantar abokin ciniki na iya aiwatar da hulɗar magana a mafi girman ma'auni.
Makomar Diarization na Kakakin Majalisa
Diarization yana haɗuwa tare da rubutawa cikin ƙira ɗaya waɗanda ke fitar da kalmomi tare da lakabin lasifika a cikin wucewa ɗaya, yana rage tara kurakurai. Yi tsammanin ingantacciyar sarrafa magana mai ma'ana, manyan tarurruka tare da mahalarta da yawa, da yawo na ainihi don taken kai tsaye. Wakilan sauti da ke kula da kai da kuma alamu na zamani (motsin lebe, alkiblar isowa daga tsararrun makirufo) za su kara kaifin daidaito, yayin da diar-diddigar na'urar zai inganta keɓaɓɓen sirri ta hanyar adana bayanan murya a gida.
Aiwatar da Gaskiyar Duniya
Ƙirƙirar bayanan tarurrukan kasuwanci masu alamar magana a cikin kayan aikin kamar Otter.ai ko Microsoft Ƙungiyoyi
Samar da "wanda ya ce menene" jerin lokuta don podcast da software na gyara hira
Fitar da rikodin cibiyar kira don raba wakili da juyi abokin ciniki don ingantaccen bincike
Ƙirƙirar ɗakin shari'a da jigon sauti ta yadda kowane mai magana ya faɗi bayaninsa daidai
Hatsari & Tsare-tsare
Rashin amfani da murya da haɗarin kwaikwaya yana ƙaruwa lokacin da aka rasa izini.
Daidaituwa na iya faɗuwa cikin lafuzza, yaruka, ko mahalli masu hayaniya.
Ana iya kuskuren sauti na roba don ingantacciyar magana ba tare da bayyananniyar lakabi ba.
Taswirar Hanya
Sami tabbataccen izini don ɗaukar murya, cloning, da sake amfani.
Gwajin ingantattun masu magana daban-daban da yanayin baya.
Ƙayyade lokacin da dole ne ɗan adam ya duba ko ya amince da abubuwan da aka fitar.
Yi lakabin sauti na roba da kuma adana bayanan da aka tabbatar don yin lissafi.
Ci gaba da Bincike
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Speaker Diarization quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Jagora na gaba
Gane Kakakin Kakakin ECAPA-TDNN
Tambayoyin da ake yawan yi
What is Speaker Diarization?
Diarization na lasifikar yana amsa tambayar "wane yayi magana yaushe?" ta hanyar rarraba rikodin sauti zuwa sassan da aka yiwa lakabi da asalin lasifikar. Yana juya rafi ɗaya na gaurayawan muryoyi zuwa jerin lokutan da ke nuna ainihin mutumin da ke magana a kowane lokaci.
Wace ainihin tambaya ce diarization na lasifikar ke son amsawa?
Diarization partitions audio by lasifika na tsawon lokaci, amsa "wanda ya yi magana lokacin" maimakon rubuta kalmomin da kansu.
Menene saka lasifika?
Ƙirƙirar lasifikar vector ce mai kayyadadden tsayi inda shirye-shiryen bidiyo daga mutum ɗaya ke kusa tare, yana ba da damar tari ta ainihi.
Wanne awo ne aka fi amfani da shi don kimanta tsarin diaryization?
DER yana haɗa maganganun da aka rasa, ƙararrawar ƙarya, da ruɗani na lasifika zuwa kashi ɗaya, yana mai da shi daidaitaccen ma'aunin diarization.
Shin daidaitaccen diarization yana buƙatar sanin ainihin sunayen masu magana a gaba?
Diarization ta tara muryoyin da sanya alamun da ba a san su ba; ba ya buƙatar sanin ainihin su wane ne mutanen da sunan su.
Me yasa ake ƙima ƙirar ƙirar jijiyoyi daga ƙarshe zuwa ƙarshen (EEND) akan bututun tari kawai?
Samfuran EEND suna amfani da horarwar da ba ta canzawa ba don yin ƙira kai tsaye ga lasifika da yawa, sarrafa magana mai jujjuyawar da ke karya tarun mai magana ɗaya-a-lokaci.