EnCodec Audio Compression
EnCodec shine babban codec na jijiya mai inganci na Meta wanda ke danne magana da kida a rahusa bitrates tare da ingantacciyar kishiya mafi nauyi.
Dubawa
It matters because it underpins modern generative audio systems and ships in open-source form for anyone to use.
Zurfafa nutsewa
An sake shi ta Meta AI a cikin 2022, EnCodec yana bin tsarin SoundStream na encoder, ragowar vector quantizer (RVQ), da na'urar tantancewa ta horar da ƙarshe zuwa ƙarshe, amma tana ƙara gyare-gyare da yawa. Yana amfani da mai rikodin juzu'i mai iya yawo, sikelin sikeli da asarar sake gina yanki na lokaci, da masu nuna wariya don ingancin fahimta. Sananniyar gudummawa ita ce ƙaramin ƙirar entropy na tushen Transformer wanda ke ƙara matsa lambobin da aka ƙididdige su ba tare da asara ba, tare da matsi ƙarin ragi ba tare da asara mai inganci ba. EnCodec kuma yana gabatar da ma'auni wanda ke daidaita yawan asarar horarwa ta atomatik don haka su tsaya tsayin daka. Yana sarrafa 24 kHz monophonic da 48 kHz audio sitiriyo, yana aiki a fadin bitrates kamar 1.5, 3, 6, da 12 kbps, kuma a 6 kbps ya kai ingancin kwatankwacin MP3 a 64 kbps. Alamun ikonta na MusicGen da AudioGen Meta.
Fahimtar Fasaha
EnCodec's encoder yana saukar da misalan tsarin igiyar ruwa tare da jujjuyawar juzu'i zuwa jeri mai ɓoye, wanda RVQ ke jujjuya su zuwa ma'auni masu tarin yawa. Samfurin yaren Transformer mara nauyi yana tsinkayar yuwuwar waɗannan alamomin da ƙididdige su, yana dawo da ƙarin matsawa kyauta. Ma'auni na horarwa yana sake daidaita gudummawar da aka bayar daga sake ginawa, na gani, da kuma asarar gaba don haka babu wani lokaci guda da ya mamaye, wanda ke kiyaye horon maƙasudi da yawa a faɗin cikakken kewayon bitrate.
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 EnCodec Audio Compression
EnCodec ya riga ya zama tsohuwar alamar alama don nau'ikan nau'ikan sauti na buɗewa da yawa, kuma zuriyarsa suna tura aminci mafi girma a ƙananan bitrates, cikakken sitiriyo da sake gina kida, da haɗa kai tare da rubutu-zuwa-audio da rubutu-zuwa- kiɗa. Yi tsammanin tallafi mai fa'ida a cikin ƙananan sadarwar bandwidth, yawo na ainihin lokaci, kuma azaman ma'auni na 'audio token' wanda ke ba da damar manyan gine-gine-samfurin-harshe su karanta da rubuta sauti.
Aiwatar da Gaskiyar Duniya
Alamar sauti don Meta's MusicGen da AudioGen janareta na rubutu-zuwa-audio
Matsa 24 kHz magana zuwa 1.5-6 kbps don watsa iyakantaccen bandwidth
Ƙirƙirar kiɗan sitiriyo 48 kHz tare da inganci kusa da MP3 a mafi girman bitrates
Yin hidima azaman codec na buɗe tushen tushen don bincike da bututun ML mai jiwuwa ta wuraren binciken da aka saki
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 EnCodec Audio Compression 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
Abun Sauti da Koyon Wakilci
Tambayoyin da ake yawan yi
What is EnCodec Audio Compression?
EnCodec shine babban codec na jijiya mai inganci na Meta wanda ke danne magana da kida a rahusa bitrates tare da ingantacciyar kishiya mafi nauyi. Yana da mahimmanci saboda yana ƙarfafa tsarin sauti na zamani da jiragen ruwa a cikin buɗaɗɗen tushe don kowa ya yi amfani da shi.
Wace kungiya ce ta fitar da EnCodec?
Meta AI (FAIR) ne ya gabatar da EnCodec a cikin 2022 a matsayin babban codec audio na jijiya.
Wane ƙarin abu ne EnCodec ya ƙara don matse ƙarin matsawa daga alamun sa?
EnCodec yana horar da ƙaramin Transformer don hasashen yiwuwar alamar alama da ƙididdiga-lambobin su, samun ƙarin matsi mara asara a saman RVQ.
A kusan 6 kbps, an ba da rahoton ingancin EnCodec a matsayin kwatankwacin MP3 akan wane bitrate?
EnCodec a 6 kbps ya kai ga inganci mai kama da MP3 a 64 kbps, babban riba mai inganci.
Wace matsala EnCodec's 'ma'auni' warwarewa yayin horo?
Tare da asara da yawa (sake ginawa, ban mamaki, abokan gaba), ma'auni yana haɓaka gradients don haka babu wata manufa ɗaya ta mamaye, ƙarfafa horo.
Wanne ainihin hanyar ƙididdigewa EnCodec ke rabawa tare da SoundStream?
Kamar SoundStream, EnCodec yana amfani da saura vector ƙididdigewa don ɓatar da abubuwan da aka haɗa su a cikin fihirisar codebook.