I-Apple PARS: Le yindlela iApple efuna ukufunda ngayo umsebenzi wakho wobuchopho endlebeni yakho.

Ingcamango yokuba ii-headphones ziyakwazi Mamela umculo wakho kwaye, kwangaxeshanye, "mamela" ingqondo yakho Kubonakala ngathi yinto ephuma kwintsomi yesayensi, kodwa i-Apple sele ivula indlela ngendibaniselwano enamandla kakhulu: izinzwa ezintsha zendlebe kunye neemodeli eziphambili ze ingqiqoEmva kwayo yonke le nto kukho indlela ebizwa I-PARS (I-Pairwise Relative Shift), indlela yokuzifundela evumela ukuba i-algorithm iqonde umsebenzi wombane wengqondo ngaphandle kokuthembela kwiingcali zokurekhoda idatha ngesandla.

Endaweni yokugxila kwigajethi ethile, uphando lwe-Apple lujolise kwindlela imodeli ye IA unakho funda ubume bexeshana lweempawu ze-EEG (i-electroencephalography) uze usebenzise olo lwazi kwimisebenzi efana nokuhlela amanqanaba obuthongo okanye ukufumanisa ukungahambi kakuhle kwemithambo-luvo. Kwaye, nangona uphononongo lungakhankanyi ngokuthe ngqo ii-AirPods, yongeza kumalungelo awodwa omenzi wechiza kunye neeprototypes ezikhomba kwikamva apho ii-headphones ezilula zinokuba luhlobo "lwelabhoratri encinci" yokujongwa kwengqondo endlebeni.

Yintoni i-Apple PARS (i-Pairwise Relative Shift) kwaye kutheni ibalulekile?

Indlela ye-PARS yavela kuphando oluthiwe thaca liqela le-Apple kunye nabadibanisi bezemfundo, kwiphepha elamkelweyo Iimodeli zeSiseko ze-NeurIPS 2025 zocweyo lweBrain kunye noMzimbaUphononongo, olunesihloko esithi "Ukufunda ukwakheka okuzalanayo kweempawu ze-EEG usebenzisa i-pairwise isalamane shift pretraining," iphakamisa indlela eyahlukileyo yokuqeqesha iimodeli ezinemiqondiso ye-electroencephalography ngaphandle kokusebenzisa iilebhile zabantu.

Ngokwesiqhelo, i-PARS bubuchule bo ukufunda ngokuzilawula isetyenziswe kwi-EEG. Esikhundleni sokucela izazi ze-neurologists ukuba zibonise ngesandla ukuba yeyiphi icandelo lomqondiso elihambelana nesigaba ngasinye sokulala okanye ukuqala kwesifo sokuxhuzula, imodeli iqeqeshwa ngedatha engachazwanga kwaye inyanzelekile ukuba isombulule ingxaki eyenziweyo kodwa eluncedo kakhulu: qikelela ukuba loluphi ixesha umgama owahlula amaqhekeza amabini omqondiso.

Uluvo olusisiseko kukuba ukuba imodeli ifunda ukuqikelela ukuba lingakanani ixesha phakathi kweefestile ezimbini ze-EEG, ekugqibeleni iyayiqonda ubume behlabathi kunye nokuxhomekeka kuluhlu olude yokusebenza kwengqondo. Oku kuthi emva koko kuvumeleke ukuba yenze ngcono kwimisebenzi yokwenyani yeklinikhi, efana nokufumanisa iipateni zokulala, ukuchonga isifo sokuxhuzula, okanye ukuqonda imiqondiso yemoto.

Ababhali bagxininisa ukuba, ngokungafaniyo neendlela zokufunda ze-EEG zakudala, ezigxininisa ngokukodwa Yakha ngokutsha iindawo ezigqunyiweyo zophawu (njengee-autoencoders ezifihliweyo, i-MAE, yenza), i-PARS igxile ekubunjweni kwexesha elihambelanayo. Oko kukuthi, ayigcwalisi "izithuba" zasekhaya, kodwa ibamba indlela amaqhekeza ahlukeneyo omqondiso adibana ngayo ngokuhamba kwexesha.

Kuvavanyo olwenziweyo, iimodeli ezisekwe kwi-PARS zibonisa ukuba ziyakwazi ukutshatisa okanye ukodlula amacebo angaphambili kwiibenchmarks ezininzi ze-EEG, ngakumbi xa iilebhile ezimbalwa zikhoyo (imeko exhaphake kakhulu kwezamayeza). Oku kwenza i-PARS ibe lukhetho olunomtsalane kuyo nayiphi na inkqubo efuna ukuthatha ithuba lemithamo emikhulu yemiqondiso yobuchopho ngaphandle kokuxhomekeka kwiingcaciso ezipheleleyo.

Isebenza njani i-PARS: ukusuka kwithokheni ukuya kuqikelelo lokufuduswa kwesihlobo

Ukusebenzisa indlela ye-PARS kwi-EEG, abaphandi bayila i-architecture esekelwe Iziguquli ezinamanyathelo amaninzi angundoqoYonke le nto iqala ngokulungiselela isiginali kwaye uyiguqule ibe ngumboniso onokuphathwa ngokulula yimodeli.

Okokuqala, uphawu lwe-EEG lwahlulwe lube iifestile zexeshana okanye "iithokheni"Le nkqubo ye-tokenization ivumela iqhekeza ngalinye lomqondiso ukuba limelwe njengeyunithi apho i-transformer inokusebenza kuyo, kufana nendlela isicatshulwa okanye imifanekiso esetyenziswa ngayo. Ezi tokens zongezwa... Ufakelo lwendawo, nangona ngendlela ethile, kuba i-PARS idlala ngokuchanekileyo ngemaski kunye nokuguqulwa kwezi zikhundla zexeshana.

Elinye lamacandelo ahlukileyo endlela kukusetyenziswa kwe ukufakela indawo okugqunyiweyoEndaweni yokunika imodeli ulwazi lwendawo ngqo kwi ixesha Kwithokheni nganye, idatha ethile yendawo ifihliwe okanye iguqulwe. Oku kunyanzela isifaki khowudi ukuba Thelekelela ubume bexeshana kumxholongaphandle kokuxhomekeka kuphela kwisalathiso okanye isitampu sexesha esicacileyo.

Intliziyo ye-PARS pre-workout ngumsebenzi we uqikelelo lweshifti yesalamane ngababiniImodeli ifumana iifestile ezimbini ze-EEG ezikhutshwe ngokungenamkhethe kwirekhodi efanayo kwaye kufuneka ziqikelele umgama wexeshana phakathi kwazo. Ayikokuqikelela nje ukuba basondelelene okanye baqelelene, kodwa malunga nokufunda imephu eqhubekayo okanye ecalulwayo ebonisa ixesha lokuphumla.

Ngenxa yoko, a idikhowuda eneendlela zokuthathela ingqaleloEli candelo libhekisa kulwazi olukhowudiweyo kuzo zombini iifestile kwaye lifunda ukunxulumana neempawu zabo zangaphakathi ukufumanisa ixesha lokuzahlula. Ndiyabulela kule nkqubo, i-transformer igqiba ukwenza imodeli yokuxhomekeka kwexesha elide kunye neepateni zokuziphendukela kwengqondo eziye zanda ngaphaya kokusingqongileyo kwendawo yee-milliseconds ezimbalwa.

Kwizigaba zamva, imodeli ilungiselelwe imisebenzi eyahlukeneyo ngokusebenzisa ukulungiswa kakuhle kwamajelo amaninzi kunye novavanyo oluthileOku kuthetha ukuba, xa sele uqeqeshwe kwangaphambili kunye ne-PARS kwiintlobo ezahlukeneyo zokurekhoda kwe-EEG (kuquka ukuseta i-electrode emininzi), ilungelelaniswe kakuhle kwimisebenzi ethile efana ukuhlelwa kokulala, ukufumanisa i-EEG engaqhelekanga okanye i-epileptic seizures.

Inqaku lobugcisa likwachaza imiba ephathekayo efana ne iiseti zedatha ezisetyenzisiweyo, ulwakhiwo oluchanekileyo lwe-encoder, uhlobo lwedekhowuda ekhethiweyo, isigqubuthelo sobuso kunye neziqwenga zesampulu yezicwangciso., Kunye ne- izixhobo zekhompyutha ezisetyenzisiweyoUkongeza, ulwahlulo loyilo olwahlukileyo luthelekiswa nezifundo zokukhupha imali ukumisela ukuba zeziphi izigqibo zoyilo ezivelisa owona msebenzi ubalaseleyo.

Ukuthelekisa nezinye iindlela: MAE, MP3, DropPos kunye nabanye

Uphononongo aluchazi nje i-PARS, kodwa luyayithelekisa nayo Iindlela zokubhekisela ekuzifundeleni kwi-EEGPhakathi kweendlela ezithelekisiweyo zi ii-autoencoders ezifihliweyo (MAE), MP3 kunye neDropPos, nganye inefilosofi eyahlukileyo xa kufikwa ekufundeni kwidatha engabhalwanga.

I-MAEs igxile kwi Yakha ngokutsha iindawo ezigqunyiweyo zophawuNgethuba lokuqeqeshwa kwangaphambili, iinxalenye zegalelo zifihliwe, kwaye imodeli izama ukuyibuyisela kwimeko. Oku kunyanzela isifaki khowudi ukuba sifunde ukumelwa okunentsingiselo, kodwa sijolise kakhulu kwiipateni zengingqi-oko kukuthi, ukuzalisa "izithuba" ezikufutshane kunokuqonda ubudlelwane bexesha elide.

IMP3 kunye nezinye iindlela ezifanayo nazo ziyaphonononga amaqhinga okubamba ingcaciso yesakhiwoNangona kunjalo, ngokweziphumo zephepha, zisasebenza ngaphantsi kune-PARS ekuboniseni ixesha elihambelanayo phakathi kwamacandelo akude ophawu.

I-DropPos, ngenxalenye yayo, ilungisa okanye isusa ulwazi olucacileyo lwendawo kwi-transformer ngengcamango yenza imodeli yomelele ngakumbi kwindawo echanekileyoNangona olu hlobo lobuchule lunceda uthungelwano ukuba lungaxhomekeki ngokugqithisileyo kuthungelwano lwendawo, kubonisiwe ukuba, ngokwalo, akwanelanga ukuxhaphaza ngokufanelekileyo ubume bexeshana beempawu ze-EEG.

Iimvavanyo zovavanyo zibonisa ukuba iimodeli iqeqeshelwe kwangaphambili nge-PARS ilingana okanye yodlula ezi zikhetho kwiibenchmarks ezine zeEEG ezisetyenzisiweyo. Apho kukhanya kakhulu kwimiba ye ukusebenza kakuhle kweleyibhileOko kukuthi, xa kuphela iqhezu lokurekhoda ezikhoyo. Oku kubalulekile kwimeko yeklinikhi, apho ukubhalwa ngokuchanekileyo kwemizuzu ye-EEG kudla ixesha elininzi kwaye kufuna iingcali ezikhethekileyo.

Isihlomelo somsebenzi sichaza ngokweenkcukacha uqwalaselo lwesiseko ngasinye, iihyperparameters ezivavanyiweyo, isiphumo samanqanaba ahlukeneyo emaski okanye ulwakhiwo lwedekhowuda eyahlukeneyo. kunye neziphumo zobungakanani bokugqibela. Umyalezo ocacileyo ngowokuba, kwi-EEG, ukufunda ngokuthe gca unxulumano lwexeshana phakathi kwamaqhekeza omqondiso kubonelela ngenzuzo ebonakalayo phezu kokwakha ngokutsha okanye ukugquma ubuso.

Iiseti zedatha ezisetyenzisiweyo: ukusuka endlebeni-EEG ebuthongweni ukuya ekubhaqweni kokuxhuzula

Ukuqinisekisa i-PARS, iApple kunye nabasebenzisana nayo basetyenzisiwe ezine ezaziwayo-kakuhle zedatha ye-EEGezibandakanya iimeko ezahlukeneyo zokusetyenziswa: ukulala, iipathologies, umsebenzi wemoto kunye nokumiselwa kwe-electrode endlebeni.

Iseti yedatha yokuqala yile Ukunxitywa kwe-Sleep Staging (EESM17), igxininise esweni lokulala kunye nezixhobo ezinxitywayo. Kubandakanya ukurekhoda ebusuku kwezifundo ze-9 kunye nenkqubo yokubeka iliso lokulala I-12-channel ear-EEG kunye ne-6-channel scalp EEGLe datha inomdla ngakumbi kuba ibonisa ukuba i-electrode ebekwe endlebeni inokubamba inxalenye ebalulekileyo yomsebenzi wobuchopho ohambelana nokwahlula izigaba zokulala.

Okwesibini ngu I-TUAB (iYunivesithi yeTempile engaqhelekanga ye-EEG Corpus), ikhoposi eyenzelwe i ukufunyanwa kwe-EEG engaqhelekangaIqokelela iirekhodi ezibhalwe njengesiqhelo okanye i-pathological, iluncedo kwiimodeli zoqeqesho ezifumanisa ukuguqulwa kwe-neurological jikelele, ngaphaya kwemeko ethile.

Owesithathu, I-TUSZ (i-Temple University Seizure Corpus), igxile kwi ukufunyanwa kokuxhuzulaIbandakanya izichasiselo eziphawula ukuqala kunye nokuphela kokuxhuzula, kunye namacandelo e-interictal. Yenye yeeseti zedatha yereferensi kwisifo sokuxhuzula sokuvavanya i-AI algorithms.

Ekugqibeleni, iseti yedatha yesine IPhysioNet-MI, kugxilwe kwi imisebenzi yengqondo yengqondoKule meko, abathathi-nxaxheba bacinga ukunyakaza (umzekelo, ukuhambisa isandla) ngelixa i-EEG irekhodwa, ivumela ukuqeqeshwa kweemodeli eziqaphela iipatheni ezinxulumene nenjongo yokunyakaza, into ephambili kwi-brain-machine interfaces.

Uqeqesho lwangaphambili lwe-PARS lwenziwa kwezi kunye nezinye iiseti zedatha ezichazwe kwi-appendix yobugcisa, ngelixa ukulungiswa kakuhle kulungiselelwe imisebenzi ethile ngaphakathi kwendawo nganye. Ukhetho lwebenchmarks ezohlukeneyo lubonisa indlela yokujonga Ayiphelelanga kwimeko yokusetyenziswa enye kwaye oko kunokusebenza njengesiseko jikelele semifuziselo ye-EEG yokuzijonga.

I-Ear-EEG kunye nee-AirPods: Uphando lwe-PARS lunxibelelana njani nee-headphone ze-Apple

Enye inkalo ephawulekayo yesi sihloko siphela kukusetyenziswa kwe indlebe-EEG, oko kukuthi, ukubanjwa kwemiqondiso yobuchopho endlebeniI-dataset ye-EESM17 sele isebenzisa iinkqubo ezibeka i-electrodes kwi-canal indlebe kunye nendlebe, endaweni ye-scalp, eyanciphisa kakhulu impembelelo ebonakalayo kunye nokuphucula induduzo.

Ngeli xesha, amaxwebhu oluntu kunye namalungelo awodwa omenzi we-Apple abonisa ukuba inkampani ibiphonononga kangangexesha elithile ii-headphones ezikwaziyo ukulinganisa i-biosignals ukusuka endlebeni. Kwisicelo selungelo elilodwa lomenzi ka-2023, inkampani ichaza “isixhobo sombane esinxitywayo” esenzelwe ukurekhoda umsebenzi wobuchopho kusetyenziswa. i-electrodes ebekwe ngaphakathi okanye ejikeleze indlebe, njengenye indlela engabonakaliyo kwiinkqubo ze-EEG ze-scalp zakudala.

Ilungelo elilodwa lomenzi wechiza ngokwalo liyavuma ukuba izisombululo ze-EEG eziqhelekileyo zifuna izixhobo ezilungiselelwe umsebenzisi ngamnye (ilungele ubungakanani kunye nokumila kwendlebe yakho, umsele wendlebe, njl. njl.), ebiza kakhulu kwaye ayinakwenzeka. Ngapha koko, nesixhobo esenziwe ngokwezifiso sinokuphulukana nokunxibelelana nolusu ngokuhamba kwexesha, sithomalalisa umgangatho wophawu.

Ukujongana nale mingeni, iApple iphakamisa isisombululo kolo xwebhu lusekwe ekubekeni ii-electrodes ezininzi kunokuba ziyimfuneko ngokungqongqo, zisasazwa kuyo yonke indlebekwaye vumela imodeli ye-AI imisele ukuba yeyiphi enika olona fundo lubalaseleyo ngalo naliphi na ixesha. Ukwenza oku, iimetrics ezifana ne I-impedance, inqanaba lengxolo, umgangatho woqhagamshelwano lwesikhumba, okanye umgama phakathi kwee-electrodes ezisebenzayo kunye nereferensi.

Nje ukuba ezi metrics zibaliwe, inkqubo iyaba ubunzima obuhlukeneyo kwi-electrode nganye kwaye idibanisa imiqondiso yayo ibe yindlela enye ephuculweyo. Ilungelo elilodwa lomenzi wechiza liquka izijekulo ezilula, ezinjengokucofa okanye ukucofa indlebe, ukuqala okanye ukumisa umlinganiselo, kunye noyilo olwahlukileyo kunye nokwahluka kokuhlanganisa okuya kwenza hardware.

Ukusuka kwithiyori ukuya kwimveliso: izinzwa kwii-AirPods kunye nokujongwa kwengqondo yemihla ngemihla

Indibaniselwano yalo mgca wophando kunye ne-PARS yenza kube lula ukucinga Ii-AirPods ezinezinzwa ezikwaziyo ukulinganisa i-EEG ukusuka kumjelo wendlebeNgapha koko, sele kukho inkqubela phambili kweli cala: i-AirPods Pro 3 ifake i-photoplethysmograph (PPG) sensor ukulinganisa izinga lentliziyo, kwaye i-Apple ibisongeza iimpawu zempilo kwizinto ezinxitywayo kule minyaka imbalwa idlulileyo.

Ukuba songeza kule electrode indlebe-EEG kunye imodeli self-esweni efana PARS ekwaziyo Ukutolika imiqondiso ngaphandle kobukhulu yolwazi kuphawuliweIsiphumo siya kuba sisixhobo esinokubona amanqanaba obuthongo, utshintsho kwingqwalasela, okanye iimpawu zokuqala zezifo ezithile zemithambo-luvo, zonke ziselubala kumsebenzisi.

Kwiimvavanyo ezichazwe kwiphepha lophando, i-algorithm ye-PARS ithatha amacandelo random zophawu lobuchopho kwaye ufunde ukuqikelela umgama wethutyana phakathi kwaboNgokusekwe kobu buchule, imodeli iphuhlisa ukuqonda okutyetyisiweyo malunga nendlela umsebenzi wobuchopho oguquka ngayo ngokuhamba kwexesha, okukhokelela kwiziphumo ezingcono xa uhlela amanqanaba okulala, ukufumana iziganeko zokuxhuzula, okanye ukwahlula okuqhelekileyo kwii-EEGs ezingaqhelekanga.

Isibheno esikhulu sale ndlela kukuba inokusebenza kwimeko apho iilebhile zinqabile. Kwimveliso yorhwebo, oko kunokuthetha ukuba ii-AirPods ezixhotyiswe ngobu buchwepheshe ziya kuba nakho ziqhelanisa nomsebenzisi ngamnye ngolwazi oluncinci oluphantsi kweliso, ukuthatha inzuzo yezixa ezikhulu zesignali ekrwada ebanjwe ngexesha lokusetyenziswa kwemihla ngemihla.

Ngokweziphumo zophononongo, iimodeli eziqeqeshwe kwangaphambili kunye ne-PARS ziyafikelela ukutshatisa okanye ukugqithisa ukuchaneka kweendlela zangaphambili kwimisebenzi eyahlukeneyo, kwaye oku kuvula umnyango wezixhobo ezibalaseleyo zabathengi, njengee-headphones, ukuqalisa ukunika imilinganiselo ebigcinelwe izixhobo zesibhedlele ezinkulu nezikhethekileyo.

Ngokuqinisekileyo, konke oku kuza nemibuzo efanelekileyo malunga ubumfihlo bedatha kunye nokhuseleko kunye nemida yokuziphatha. Ingcinga yokuba i-AI inokwazi kungekuphela nje izinga lentliziyo yakho okanye amanyathelo akho emihla ngemihla, kodwa kunye neepateni zakho zokusebenza kwengqondo, ivelisa inani elithile lenkxalabo. Okwangoku, zombini iphepha kunye namalungelo awodwa omenzi wechiza ahlala kwindawo yophando kunye noyilo lwengqikelelo, ngaphandle komhla othile wemveliso yorhwebo.

Ukusetyenziswa okunokwenzeka: impilo, impilo, ukuqhuba, kunye nokusebenza kwengqondo

Ukuba i-PARS kunye ne-ear-EEG-based technology ibamba i-AirPods ezizayo okanye ezinye izinto ezinxitywayo, uluhlu lwezicelo lunokuba lubanzi kakhulu. Okokuqala, iya kuba yinto efanelekileyo ukubeka iliso ukulala ngokuqhubekayo nangokukhululekileyo, ihlela ngokuzenzekelayo izigaba ze-REM ne-NREM (i-NREM 1, i-NREM 2 kunye ne-NREM 3) kunye nokubonelela ngeenkcukacha ezicacileyo kumgangatho wokuphumla.

Ngaphezu koko, ukufunyanwa kwe amanqanaba okuhoya, iziqendu zoxinzelelo, okanye iimeko zokuphaphama Oku kunokuba luncedo kakhulu kwiimeko ezinjengokuqhuba, umsebenzi onzima ngengqondo, okanye ukufunda. Isixhobo esichaza ukwehla ngequbuliso kwingqalelo sinokulumkisa umsebenzisi xa esengozini yokulala eqhuba okanye enze iimpazamo ezinzulu ngenxa yokudinwa.

Kwimeko yeklinikhi, ukubeka iliso kwengqondo okulumkileyo kunye nokuqhubekayo kunokuququzelela ukufunyaniswa kwangoko kweziphazamiso ezifana nokuxhuzula, iingxaki zokulala, okanye izifo ze-neurodeergenerativeAyikokutshintsha i-neurologist okanye isibhedlele, kodwa malunga nokubonelela ngedatha ebalulekileyo enokusebenza njengesilumkiso kwangethuba okanye incedise ekuxilongweni.

Omnye umgca onomdla ngu I-biofeedback igxile kwimpilo-ntle yengqondoUkuba isixhobo siyakwazi ukunxulumanisa iipatheni ezithile ze-EEG neemeko zokuphumla, ukugxininiswa okunzulu, okanye uxinzelelo, sinokukhokela umsebenzisi kuqheliselo lokuphefumla, ukucamngca, okanye uqeqesho lokuqonda, lubonelela ngezalathi zexesha lokwenyani zokuba ngaba ezi nkqubo zinesiphumo esinqwenelekayo.

Ibingayi kuma nje engqondweni. Amaxwebhu oluntu ka-Apple akhankanya ukongezwa okunokwenzeka kwe izinzwa zokulinganisa umthamo wegazi, ukusebenza kwezihlunu zobuso, kunye neentshukumo zamehlo ukusuka kwi-earphone ngokwayo. Idityaniswe nomqondiso we-EEG kwaye iqhutywe kwi-a iPhone okanye esinye isixhobo, le datha inokondla imifuziselo ye-AI ekwaziyo ukunika umfanekiso opheleleyo wesimo somzimba kunye neemvakalelo zomsebenzisi.

Konke oku kusetyenziswa kuya kufuneka kukhatshwe ulawulo olungqongqo kwimvume, ulawulo lwedatha kunye nokufikelela ngamaqela esithathuKwimizekelo enikeziweyo, kucingelwa ukuba ulwazi lunokwabelwana nabasebenzi bezempilo kuphela ukuba umsebenzisi uyalugunyazisa, kwaye uninzi lokucubungula lwenziwa ekuhlaleni ukunciphisa umngcipheko.

Ngokwenyani, yintoni eboniswa liphepha le-PARS kunye namalungelo awodwa omenzi wechiza kunye neeprototypes kukudibana okucacileyo: iApple iyaphonononga, kwelinye icala, eyona ndlela yokuqokelela imiqondiso yobuchopho endlebeni nakwelinye eyona ndlela ingcono yokutolika nge-AI ngaphandle kokuxhomekeka kwiinkcazo zabantuUkuba zombini iziqwenga ziyadibana, ii-headphones zinokuyeka ukuba ngabadlali beaudio ezilula kwaye zibe zizixhobo zempilo eziphucukileyo kunye nokusebenza, ngaphandle kokuba imiba yokuziphatha kunye nemfihlo ilawulwa kakuhle.

Yonke into ikhomba kwinto yokuba sisekuqaleni kwesizukulwana esitsha sokunxiba apho Iindlela ezifana ne-Apple PARS (Parwise Relative Shift) kunye ne-discrete sensors njenge-ear-EEG Banokuguqula indlela esikuqonda ngayo ukulala kwethu, ingqalelo yethu, kunye nempilo yethu yemithambo-luvo, ukuqala ngento yemihla ngemihla njengee-AirPods.

Inqaku elidibeneyo:

Yintoni uMfundi weSemantic: i-injini yokukhangela yezemfundo ene-AI

UIsake

Umbhali onomdla malunga nehlabathi le-bytes kunye netekhnoloji ngokubanzi. Ndiyakuthanda ukwabelana ngolwazi lwam ngokubhala, kwaye yile nto ndiza kuyenza kule bhlog, ndikubonise zonke izinto ezinomdla malunga nezixhobo, isoftware, ihardware, iindlela zetekhnoloji, kunye nokunye. Injongo yam kukukunceda uhambe kwihlabathi ledijithali ngendlela elula neyonwabisayo.