Saurari Halarta da Tafsiri
Saurara, Halartar da Harafi (LAS) alama ce ta hanyar sadarwa ta jijiya ta 2015 wacce ke kwafin magana kai tsaye zuwa haruffa, ba tare da ƙamus ɗin da aka gina da hannu ba ko ƙirar harshe daban.
Dubawa
It showed that a single end-to-end model could do speech recognition.
Zurfafa nutsewa
Saurari, Halarci da Harafi, wanda Google masu bincike Chan, Jaitly, Le, da Vinyals suka gabatar a cikin 2015, ya kasance ɗaya daga cikin farkon masu gane magana daga ƙarshe zuwa ƙarshe. Yana da sassa biyu: 'Mai sauraro,' LSTM na pyramidal bidirectional wanda ke ɓoye sautin yayin da yake rage girman lokaci, da kuma 'Speller,' mai ƙirar LSTM na tushen hankali wanda ke fitar da haruffa ɗaya bayan ɗaya. Tsarin kulawa yana barin Mai Tafsiri ya mai da hankali kan yanki na jiwuwa mai dacewa ga kowane harafin fitarwa. Ba kamar tsofaffin bututun HMM-DNN ba, LAS ba ta buƙatar ƙamus na sauti, babu daidaitawar tilastawa, kuma babu ƙirar harshe dabam dabam; yana koyan harrufa, iyakoki na kalmomi, da acoustics tare daga sautin da aka rubuta. Kai tsaye ya yi wahayi zuwa jeri-zuwa-jeri na zamani da tsarin ASR na tushen hankali.
Fahimtar Fasaha
LAS yana haɗa mai rikodin-dikodi tare da hankali. Pyramidal LSTM encoder yana rage ƙudurin lokaci a kowane yadudduka uku, yana yanke dogon jerin sauti zuwa tsayin da za a iya sarrafawa don haka hankali yana iya faɗowa. A kowane mataki na yanke hukunci mai Speller yana ƙididdige ma'aunin hankali a kan duk jahohin rikodi, ya haɗa su cikin mahallin mahallin, kuma yana tsinkayar hali na gaba. Horowa yana haɓaka yuwuwar jerin halaye daidai; dabarar da aka tsara-samfurin yana rage rashin daidaituwar jirgin kasa/gwaji.
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 Halartar Sauraro da Tafsiri
LAS yanzu tarihi ne, amma DNA ɗin sa yana gudana ta kowane tsarin ASR na zamani. Tunanin tushen sa mai rikodin rikodin sa ido ya samo asali zuwa masu gano masu Canjawa da Conformer, yayin da hanyoyin da suka danganci su kamar RNN-Transducer ikon kan na'urar. Tsare-tsare na gaba suna ci gaba da wannan yanayin ƙarshen zuwa ƙarshe, suna ba da izini tare da fassara da fahimta a cikin nau'ikan harsuna da yawa, da turawa zuwa yawo, ƙaramin latency rubutu wanda LAS, kasancewar ba yawo, ba zai iya samarwa da asali ba.
Aiwatar da Gaskiyar Duniya
Rubuta Turanci da aka faɗa kai tsaye zuwa haruffa ba tare da ƙamus na furci ba
Yin hidima a matsayin tushen ra'ayi don tushen hankali da tsarin sautin murya
Nuna horon ƙarshe zuwa ƙarshe don aikin kwas ɗin fahimtar magana-ilimi da ma'auni
Samfuran jeri-zuwa-jeri mai ban sha'awa daga baya da aka yi amfani da su a cikin bututun fassarar magana
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 Listen Attend and Spell 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
Kiɗa ta atomatik
Tambayoyin da ake yawan yi
What is Listen Attend and Spell?
Saurara, Halartar da Harafi (LAS) alama ce ta hanyar sadarwa ta jijiya ta 2015 wacce ke kwafin magana kai tsaye zuwa haruffa, ba tare da ƙamus ɗin da aka gina da hannu ba ko ƙirar harshe daban. Ya nuna cewa samfurin ƙarshe-zuwa-ƙarshen zai iya yin fahimtar magana.
Menene manyan abubuwa guda biyu na samfurin LAS?
LAS ya ƙunshi maɓalli na 'Mai sauraro' wanda ke sarrafa sauti da na'urar 'Speller' mai tushen hankali wanda ke fitar da haruffa.
Menene fitar da Speller a cikin LAS?
Mai Rubutun Ƙaƙwalwar Dikodi wanda ke samar da halayen rubutun ta hali, koyan rubutun kalmomi kai tsaye.
Me yasa ake kiran maƙallan LAS 'pyramidal'?
Kowane Layer na pyramidal BLSTM yana raba tsayin jeri, yana rage tsawon jerin sauti don haka hankali yana da yuwuwar ƙididdigewa.
Wane tsofaffin bangaren LAS musamman BAYA buƙata?
Ba kamar na bututun HMM-DNN na gargajiya ba, LAS na koyo kai tsaye daga nau'ikan sauti-zuwa-rubutu ba tare da ƙamus na waya ko daidaitawar tilastawa ba.
Wace hanya ce ke ba Mai Tafsirin damar mayar da hankali kan abin da ya dace na sauti ga kowane hali?
Tsarin kulawa yana ƙididdige ma'auni akan jahohin mai rikodin, yana samar da mahallin mahallin mahallin mahallin yana amfani da shi a kowane mataki.