UMHLAHLANDLELA WE-AI womsindo

I-PESQ kanye ne-STOI Speech Quality Metrics

I-PESQ ne-STOI zingamamethrikhi enhloso ajwayelekile abonisa ukuthi inkulumo ecutshunguliwe izwakala kahle kangakanani nokuthi iqondakala kanjani, ngaphandle kokudinga abalaleli abangabantu.

2 amaminithi ukufundaIgcine ukubuyekezwa

Uhlolojikelele

They let engineers benchmark codecs, noise reducers, and speech-enhancement models automatically.

I-Deep Dive

I-PESQ (Ukuhlola Okubonwayo Kwekhwalithi Yokukhuluma), esezingeni njenge-ITU-T P.862, ibikezela ikhwalithi ecatshangwayo yenkulumo, ikakhulukazi yocingo kanye nokuhlolwa kwekhodekhi. Iqhathanisa isignali yereferensi ehlanzekile necekelwe phansi futhi ikhipha amaphuzu esikalini esifana ne-MOS (cishe -0.5 kuya ku-4.5), ifanekisela umbono womuntu wokuzwa. I-STOI (I-Short-Time Objective Intelligibility), eyethulwe ngo-2010, esikhundleni salokho ibikezela ukuqondakala: mangaki amagama umlaleli angawaqonda ngempela. Ihlobanisa izimvilophu zesikhashana zenkulumo ehlanzekile necutshunguliwe kuwo wonke amabhendi emvamisa, ikhiqiza amaphuzu asuka ku-0 kuye ku-1. Womabili amamethrikhi angenele (asekelwe kunkomba). I-PESQ iyaphendula 'ingabe izwakala kahle?' kuyilapho u-STOI ephendula ngokuthi 'uyayiqonda?' Ehlangene angamathuluzi okuhlola azenzakalelayo okuthuthukisa inkulumo, i-denoising, namasistimu e-derverberation.

I-Technical Insight

Womabili amamethrikhi ayaphazamisa: aqondanisa isithenjwa esihlanzekile nesiginali eyehlisiwe ngaphambi kokushaya amagoli. I-PESQ yenza amamephu womabili amasiginali esikalini somsindo wengqondo (ama-Bark bands), ibala ukuphazamiseka kokubona ngokuhamba kwesikhathi, bese ikubuyisela kunani elifana ne-MOS. I-STOI ihlukanisa inkulumo ibe amabhendi e-octave eyodwa kwezintathu, ithatha amasegimenti emvilophu amafushane angu-~400 ms, iziqeshana futhi iwenze abejwayelekile, bese ibala ukuhlobana phakathi kwereferensi nezimvilophu ezonakele. Ukulinganisa lokho kuhlobana kunikeza amaphuzu okuqonda angu-0 kuye ku-1.

I-Strategic Impact

Finyelela futhi ufinyelele

Ithuthukisa ukufinyeleleka ngokuloba, ukulandisa, nezixhumi ezibonakalayo zezwi.

Izindleko kanye nesabelomali

Amaqembu emidiya angathumela umsindo opholishiwe ngokushesha ngamabhajethi amancane.

Isivinini nesikali

Amasistimu abhekene nekhasimende angacubungula ukusebenzelana okukhulunyiwe ngesilinganiso esikhulu.

Ikusasa le-PESQ kanye ne-STOI Speech Quality Metrics

Ngenxa yokuthi i-PESQ ne-STOI zidinga ireferensi ehlanzekile, ucwaningo lushintshela kumamethrikhi angaphazamisi, angenazithenjwa afana ne-DNSMOS ne-NISQA athola ikhwalithi kusuka kusiginali eyonakele yodwa kusetshenziswa amanethiwekhi emizwa. Amamodeli amasha okufunda ngokujulile nawo aqeqeshelwe ukubikezela i-MOS yomuntu ngokuqondile. Noma kunjalo, i-PESQ kanye ne-STOI kusalokhu kuyizimpawu zokubonisa ezigxilile, futhi inkambiso eyinhloko izenza zihlukaniseke ukuze zisetshenziswe ngokuqondile njengemisebenzi yokulahlekelwa kokuqeqeshwa kwamanethiwekhi athuthukisa inkulumo esikhundleni sokuhlola kuphela ngemva kweqiniso.

Ukuqaliswa Komhlaba Wangempela

Izilinganiso zokuthuthukisa inkulumo kanye namamodeli okunciphisa umsindo kumasethi okuhlola ajwayelekile

Ukuqhathanisa ifoni nekhwalithi ye-codec ye-VoIP ngesikhathi sobunjiniyela benethiwekhi

Ukushuna insiza yokuzwa kanye nokucubungula kwe-cochlear-implant ukuze kucace kakhulu

Ukuqinisekisa ama-algorithms e-derverberation ezinkomfeni namapayipi omsizi wezwi

Izingozi & Guardrails

Ukusetshenziswa kabi kwezwi kanye nezingozi zokuzenza ongeyena ziyanda uma imvume ingekho.

Ukunemba kungase kwehle kuzo zonke izinhlobo zokuphimisela, izilimi zesigodi, noma izindawo ezinomsindo.

Umsindo wokwenziwa ungenziwa iphutha njengenkulumo eyiqiniso ngaphandle kokulebula okucacile.

Ukuqalisa Umhlahlandlela

1

Thola imvume esobala yokuthwebula izwi, ukuhlanganisa, nokusebenzisa kabusha.

2

Ikhwalithi yokuhlola kuzo zonke izipikha nezimo zangemuva.

3

Chaza ukuthi kunini lapho umuntu kufanele abuyekeze noma agunyaze okuphumayo.

4

Lebula umsindo wokwenziwa futhi ugcine amarekhodi atholakalayo ukuze aziphendulele.

Qhubeka Uhlole

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 PESQ and STOI Speech Quality Metrics quiz

Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.

Qala imibuzo

Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation

Umhlahlandlela olandelayo

Ikhwalithi Yokwenziwa Kwenkulumo

Imibuzo evame ukubuzwa

What is PESQ and STOI Speech Quality Metrics?

I-PESQ ne-STOI zingamamethrikhi enhloso ajwayelekile abonisa ukuthi inkulumo ecutshunguliwe izwakala kahle kangakanani nokuthi iqondakala kanjani, ngaphandle kokudinga abalaleli abangabantu. Bavumela onjiniyela be-benchmark codec, izinciphisi zomsindo, namamodeli athuthukisa inkulumo ngokuzenzakalelayo.

I-PESQ ikala ini ngokuyinhloko?

I-PESQ (ITU-T P.862) ibikezela ikhwalithi yenkulumo ecatshangwayo esikalini esifana ne-MOS.

I-STOI ibikezela ini ngokuyinhloko?

I-STOI ilinganisela ukuqondakala, okungukuthi indlela inkulumo eqondakala ngayo, esikalini sika-0 kuya ku-1.

Kokubili i-PESQ ne-STOI kuchazwa njengamamethrikhi 'aphazamisayo'. Kusho ukuthini lokho?

Amamethrikhi angenele aqhathanisa isiginali eyehlisiwe ngokumelene nereferensi ehlanzekile ukuze kubalwe amaphuzu.

Ingakanani ibanga lemiphumela ye-STOI?

I-STOI ikhipha inani eliphakathi kuka-0 no-1, lapho ukuphakama kusho okuqondakala kakhulu.

I-PESQ imodeli yokuzwa komuntu isebenzisa luphi uhlobo lwesikali sokubona?

I-PESQ ibeka amamephu amasignali emelela umsindo we-psychoacoustic usebenzisa i-Bark-band processing.