Iphepha Elimhlophe: Izinzuzo ze-Edge AI + Cloud-based Security Systems

I-Edge Computing + AI = I-Edge AI

• I-AI ku-Smart Security Terminals
• I-Edge AI ku-Access Control
• I-Edge AI ekugadweni kwevidiyo

 

I-Cloud Platform Ye-Edge Data Storage and Processing iyadingeka

• Isistimu Yokulawula Ukufinyelela Okusekelwe Efwini
• Uhlelo Lokubhekwa Kwevidiyo olusekelwe efwini
• Izinzuzo Zesistimu Yokuphepha Esekelwe Efu Yesihlanganisi Sesixazululo Nesifaki

 

Izinselelo Ezivamile Ibhizinisi Lanamuhla libhekana nazo ekufakeni inkundla ye-Edge AI + Cloud kusixazululo se-Video Surveillance

• Ikhambi

 

• Isendlalelo

Ukuthuthuka kwakamuva kwezobuchwepheshe kwenze kwaba lula ukunciphisa ubungozi nokuvikela indawo yakho yokusebenza. Amabhizinisi amaningi amukele ukuqamba okusha futhi athola izixazululo zokuphatha isikhathi sabasebenzi kanye nezinkinga zokuphatha isikhala. Ikakhulukazi emabhizinisini amancane esimanje, ukuba nohlelo olufanele lokuvikela oluhlakaniphile kungenza umehluko ekugcineni indawo yakho yokusebenza, kanye nempahla yakho, iphephile. Futhi, kuyasiza ukulawula nokwenza ngcono isevisi yamakhasimende, nokuqapha ukusebenza kwabasebenzi.

Ukulawula Ukufinyelela & Surveillance Video izingxenye ezimbili ezibalulekile zokuvikeleka okuhlakaniphile. Abantu abaningi manje sebejwayele ukungena ehhovisi besebenzisa ukubona ubuso nokuhlola ukuphepha kwendawo yokusebenza ngokugadwa kwevidiyo.

Ngokombiko we-ResearchAndMarkets.com, iGlobal Video Surveillance Market ilinganiselwa ku-USD 42.7 Bn ngo-2021 futhi kulindeleke ukuthi ifinyelele ku-USD 69.4 Bn ngo-2026, ikhula ku-CAGR ye-10.2%. I-Global Access Control Market ifinyelele inani lama-US$ 8.5 Bilion ngo-2021. Uma sibheka phambili, imakethe kulindeleke ukuthi ifinyelele ku-US$ 13.5 Bilion ngo-2027, ikhombisa ku-CAGR ka-8.01% (2022-2027).

Amabhizinisi anamuhla anethuba elingakaze libonwe lokuthola izinzuzo zezixazululo zokuphepha ezihlakaniphile. Labo abakwazi ukwamukela intuthuko entsha ekwakhiweni kwesistimu yezokuphepha bangabhekana nezingozi zokuphepha ngaso sonke isikhathi futhi bazuze izinzuzo ezinkulu ekutshalweni kwezimali kwesistimu yabo yezokuphepha. Leli phepha elimhlophe labelana ngezizathu zokuthi kungani iplatifomu esekwe e-Edge AI + Cloud kufanele ibe yisinqumo sokuqala samabhizinisi esimanje.

 

• 
  
  

• I-Edge Computing + AI = I-Edge AI

  Ngokungafani ne-Cloud computing, I-Edge computing iyisevisi yekhompuyutha ehlukaniselwe indawo ehlanganisa ukugcinwa, ukucutshungulwa, nezinhlelo zokusebenza. I-Edge ibhekisela kumaseva atholakala esifundeni futhi aseduze nezindawo zokugcina, njengamakhamera agadayo nezinzwa, lapho idatha ithathwa khona kuqala. Le ndlela inciphisa inani ledatha okufanele ihambe ngenethiwekhi ukuze kubangele ukubambezeleka okuncane. I-Edge computing kucatshangwa ukuthi ithuthukisa i-Cloud computing ngokwenza i-Data Analytics eduze ngangokunokwenzeka kumthombo wedatha.

Ekusetshenzisweni okufanele, yonke imithwalo yomsebenzi izobekwa phakathi nefu ukuze ujabulele izinzuzo zesikali nobulula kusuka ku-cloud-AI. Kodwa-ke, ukukhathazeka okuvela kumabhizinisi esimanje mayelana ne-latency, ukuphepha, umkhawulokudonsa, kanye nokuzimela kubiza ukuthunyelwa kwemodeli ye-artificial intelligence (AI) e-Edge. Kwenza izibalo eziyinkimbinkimbi ezifana ANPR noma ukutholwa okusekelwe ku-AI okuthengekayo kumakhasimende angahlosile ukuthenga iseva yendawo ye-AI eyinkimbinkimbi futhi achithe isikhathi ekuyilungiselela.

I-Edge AI empeleni iyi-AI esebenzisa i-Edge computing ukuze isebenzise idatha endaweni, ngaleyo ndlela isebenzise izinzuzo ezinikezwa yi-Edge computing. Ngamanye amazwi, ukubala kwe-AI kwenziwa kumadivayisi aseduze nomsebenzisi emaphethelweni enethiwekhi, eduze nalapho idatha itholakala khona, kunokuba phakathi nendawo endaweni yekhompyutha yamafu noma isikhungo sedatha esiyimfihlo. Amadivayisi anezinzwa ezifanele namaphrosesa, futhi awadingi uxhumano lwenethiwekhi ukuze acubungule idatha futhi athathe isinyathelo. Ngakho-ke, i-Edge AI inikeza isisombululo sokushiyeka kwe-AI encike efwini.

Abathengisi abaningi abahamba phambili bezokuphepha ngokomzimba sebevele besebenzisa i-AI enqenqemeni ekulawuleni ukufinyelela kanye nokubhekwa kwevidiyo ukuze kuthuthukiswe ukusebenza kahle nokunciphisa izindleko zizonke zokukhiqiza/isevisi. Lapha, i-AI enqenqemeni izodlala indima ebalulekile.

• I-AI ku-Smart Security Terminals

  Njengoba ama-algorithms wenethiwekhi ye-Neural kanye nengqalasizinda ye-AI ehlobene ithuthuka, i-Edge AI yethulwa ezinhlelweni zokuphepha zezentengiselwano.
  
  Amabhizinisi amaningi esimanje asebenzisa i-AI yokuqaphela into eshumekwe kumatheminali ahlakaniphile ukuze avikeleke futhi avikeleke emsebenzini. I-AI yokuqashelwa kwento ene-algorithm yenethiwekhi ye-neural eqinile ikwazi ukubona kalula izici kunoma iyiphi ividiyo noma isithombe, njengabantu, izimoto, izinto nokunye. Bese ikwazi ukuhlaziya futhi ikhiphe izici zesithombe. Isibonelo, ingathola ukuba khona kwabantu abasolisayo noma izimoto endaweni ebucayi.

• 

Ukuqashelwa kobuso kwe-Edge kuwubuchwepheshe obuthembele kukho kokubili i-Edge computing kanye ne-Edge AI, ethuthukisa ngokuphawulekayo isivinini, ukuphepha, nokuthembeka kwamadivayisi okulawula ukufinyelela. Lapho isetshenziselwa ukulawula ukufinyelela, ukubonwa kobuso kwe-Edge kuqhathanisa ubuso obuthulwe endaweni yokufinyelela kusizindalwazi sabantu abagunyaziwe ukuze kunqunywe ukuthi ukhona yini okufanayo. Uma kukhona okufanayo, ukufinyelela kunikezwa, futhi uma kungekho okufanayo, ukufinyelela kuyenqatshwa futhi isexwayiso sezokuphepha singaqaliswa.

Ukuqashelwa kobuso okuncike ku-Edge computing kanye ne-Edge AI kungacubungula idatha endaweni (ngaphandle kokuyithumela emafini). Ngenxa yokuthi idatha isengozini enkulu yokuhlaselwa ngesikhathi sokudluliselwa, ukuyigcina emthonjeni lapho ikhiqizwa khona kunciphisa kakhulu ithuba lokwebiwa kolwazi.

I-Edge AI iyakwazi ukuhlukanisa phakathi kwabantu bempilo yangempela kanye nama-spoof angaphili. Ukutholwa kokuphila ku-Edge kuvimbela ukuhlaselwa kobuso kusetshenziswa i-2D ne-3D (isithombe esimile noma esiguqukayo nevidiyo).

• 
  

• Ukwehluleka kwezobuchwepheshe okumbalwa

  Ukuqashelwa kobuso kwe-Edge kuwubuchwepheshe obuthembele kukho kokubili i-Edge computing kanye ne-Edge AI, ethuthukisa ngokuphawulekayo isivinini, ukuphepha, nokuthembeka kwamadivayisi okulawula ukufinyelela. Lapho isetshenziselwa ukulawula ukufinyelela, ukubonwa kobuso kwe-Edge kuqhathanisa ubuso obuthulwe endaweni yokufinyelela kusizindalwazi sabantu abagunyaziwe ukuze kunqunywe ukuthi ukhona yini okufanayo. Uma kukhona okufanayo, ukufinyelela kunikezwa, futhi uma kungekho okufanayo, ukufinyelela kuyenqatshwa futhi isexwayiso sezokuphepha singaqaliswa.

 

Ukuncipha kwamathuba okuntshontshwa kolwazi

Ukusebenzisa ukuqashelwa kobuso kuzixazululo zokulawula ukufinyelela nakho kuyathrenda, ikakhulukazi emhlabeni wamanje wamabhizinisi, lapho kunokukhathazeka okusabalele mayelana nokusebenza kahle kanye nezindleko. Ngenxa yalokho esikufundile phakathi nobhubhane, kunesidingo esikhulayo sokususa 'ukungqubuzana' kokuhlangenwe nakho komsebenzisi.
 

Ukutholwa kosongo okuthuthukisiwe ngokutholwa kokuphila

Ukuqashelwa kobuso I-AI eshumekwe ekulawuleni ukufinyelela kwesimanje kanye namakhamera okugada iwukusetshenziswa okuvamile kwalobu buchwepheshe kwezokuvikela.

Ihlonza izici zobuso bomuntu futhi iziguqulele ku-matrix yedatha. Lawa matrices wedatha agcinwa kumatheminali e-Edge noma ifu ukuze ahlaziywe, izinqumo zebhizinisi eziqhutshwa idatha, kanye nokuthuthukiswa kwenqubomgomo yezokuphepha.

 

• I-Edge AI ekugadweni kwevidiyo

  Empeleni, isisombululo se-Edge AI sibeka ubuchopho kuyo yonke ikhamera exhunywe nohlelo, ekwazi ukuhlaziya ngokushesha futhi idlulisele ulwazi olufanele kuphela efwini ukuze lugcinwe.
  
  Ngokuphambene nohlelo lwendabuko lokuvikela ividiyo oluhambisa yonke idatha isuka kukhamera ngayinye iye kusizindalwazi esisodwa esiphakathi ukuze ihlaziywe, i-Edge AI yenza amakhamera ahlakaniphe - ihlaziya idatha khona kanye emthonjeni(ikhamera) bese ihambisa idatha efanele nebalulekile kuphela. ifu, ngaleyo ndlela esusa izindleko ezibalulekile zamaseva edatha, umkhawulokudonsa owengeziwe, nezindleko zengqalasizinda ngokuvamile ezihlotshaniswa nokuqoqwa nokuhlaziywa kwevidiyo enevolumu ephezulu.

• 

 

Ukusetshenziswa komkhawulokudonsa ophansi

Inzuzo enkulu ye-Edge AI ukunciphisa ukusetshenziswa komkhawulokudonsa. Ezifakelweni eziningi umkhawulokudonsa wenethiwekhi uwumkhawulo ngakho-ke ividiyo icindezelwa kakhulu. Ukwenza ukuhlaziya kwevidiyo okuthuthukile kuvidiyo ecindezelwe kakhulu kunciphisa ukunemba kwezibalo, ngakho-ke ukucubungula idatha yangempela ku-Edge kunezinzuzo ezicacile.
 

Ukusabela okusheshayo

Enye inzuzo enkulu yokwenza ikhompuyutha kukhamera ukunciphisa ukubambezeleka. Kunokuba ithumele ividiyo ngemuva ukuze icutshungulwe futhi ihlaziywe, ikhamera enokubona ubuso, ukutholwa kwemoto, noma ukutholwa kwento ingabona umuntu ongafunwa noma osolisayo futhi ixwayise abasebenzi bezokuphepha ngokushesha.
 

Ukwehliswa kwezindleko zabasebenzi

Ngaleso sikhathi, ivumela abasebenzi bezokuphepha ukuthi bagxile ezintweni ezibaluleke kakhulu/ezehlakalweni. Amathuluzi afana nokutholwa kwabantu, ukutholwa kwemoto, noma ukutholwa kwento kungaxwayisa ngokuzenzakalelayo abasebenzi bokuphepha bemicimbi. Lapho kusetshenziswa khona ukuqapha okubukhoma, abasebenzi bangenza okwengeziwe ngabantu abancane ngokuhlunga okuphakelayo kwekhamera ngaphandle komsebenzi othize futhi basebenzise imibono yangokwezifiso ukuze babone izindawo ezithile kuphela noma amakhamera.

 

•I-Cloud Platform Ye-Edge Data Storage and Processing iyadingeka

Njengoba inani lokurekhodwa okuvela kumakhamera agadayo likhula nsuku zonke, ngakho-ke inkinga yokugcina izingobo zomlando zedatha enkulu kangaka iya iba ibalulekile. Enye indlela endaweni yokugcina kungaba ukudlulisa ividiyo kuplathifomu yesofthiwe esekwe emafini.

Amakhasimende manje aya ngokuya eba nesidingo ngezinhlelo zawo zokuphepha, alindele izimpendulo ezisheshayo ezinkingeni zawo. Ngaleso sikhathi, baphinde balindele ukuthi uhlelo lunezinzuzo ezijwayelekile ezihlotshaniswa nanoma yikuphi ukuguqulwa kwedijithali - ukuphathwa okumaphakathi, izisombululo ezingaka, ukufinyelela kumathuluzi adinga ukucutshungulwa okunamandla, kanye nokwehliswa kwezindleko.

Isistimu yokuvikela ebonakalayo esekelwe emafini isiba inketho ethandwayo ngokushesha njengoba kwenzeka ukuthi izinhlangano zicubungule inani elikhulu ledatha emafini ngezindleko eziphansi kanye nokusebenza kahle kokuphatha. Ngokuyisa ingqalasizinda ebizayo emafini, izinhlangano ngokuvamile zingabona ukwehla kwenani eliphelele lezindleko zokuphepha ngamaphesenti angama-20 kuye kwangama-30.

Ngokukhula okusheshayo kwe-cloud computing, imakethe nezindlela izixazululo zokuphepha eziphathwa ngayo, ezifakwe futhi ezithengiwe zishintsha ngokushesha.

• Amasistimu Wokulawula Ukufinyelela Asekelwe Efwini

Ikhonsoli eyodwa yokuphatha amasayithi amaningi

Ifu livumela izinhlangano ukuthi zilawule ukubhekwa kwevidiyo yazo kanye nokulawula ukufinyelela ezindaweni eziningi kusuka kufasitelana elilodwa lengilazi. Lokhu kwenza kube lula ukulawula amakhamera, iminyango, izexwayiso nezimvume zezakhiwo zabo, izindawo zokugcina izimpahla, kanye nezitolo zezitolo kusuka noma yikuphi emhlabeni. Njengoba idatha ingabiwa kalula ngefu, ulwazi lungafinyelelwa ngokushesha.
 

Ukuphathwa komsebenzisi okuvumelana nezimo ukuze kuthuthukiswe ukuvikeleka

Abaphathi bangahoxisa ukufinyelela nganoma yisiphi isikhathi, kunoma iyiphi indawo, banikeze ukuthula kwengqondo uma kwenzeka ibheji lilahlekile noma lintshontshwa noma esimweni esingavamile lapho isisebenzi siziphatha kabi. Ngokunjalo, abalawuli banganikeza ukufinyelela kwesikhashana ezindaweni ezivikelekile njengoba kudingeka, baqondise ukuvakasha komthengisi nosonkontileka. Amasistimu amaningi futhi afaka ukulawula kokufinyelela okusekelwe eqenjini, okunekhono lokuqoka izimvume ngomnyango noma phansi, noma ukumisa isigaba esivumela abasebenzisi abathile ezindaweni ezikhawulelwe.

• Imisebenzi enokukala

  Ukuphepha kungalinganiswa kalula ngokuhlanganisa yonke into ngefu. Inombolo engenamkhawulo yamakhamera nezindawo zokulawula ukufinyelela zingangezwa kuplathifomu yamafu. Amadeshibhodi asiza ukugcina idatha ihlelekile. Kunesixazululo saso sonke isimo njengoba ukala, njengamasango, izindawo zokupaka, izindawo zokugcina izimpahla, nezindawo ezingenakho ukufinyelela kunethiwekhi.

• 

Ukusebenziseka kalula

Uhlelo olusekelwe emafini luphinde lwenzelwe ukuthi lube lula, njengoba luvumela abasebenzi nabavakashi ukufinyelela besebenzisa amadivaysi abo eselula. Lokhu kulungele abasebenzi njengoba ukhiye wabo ungenamthungo, uyaphatheka, futhi vele unabo ngaso sonke isikhathi. Ilungele futhi amabhizinisi, njengoba egwema ubunzima nezindleko zokuphrinta “okhiye” abasha kubasebenzi nabavakashi.
 

• Izinhlelo Zokubhekwa Kwevidiyo ezisekelwe efwini

Isistimu yokuvikela ividiyo esekelwe emafini iwuhlobo lwesistimu yokuvikela erekhoda amavidiyo nge-inthanethi esikhundleni sokuwaqopha kudivayisi yokugcina indawo. Aqukethe izindawo zokugcina zekhamera yevidiyo ye-AI ezixhumeka kumhlinzeki wakho wokuphepha wamafu nge-inthanethi. Lo mhlinzeki wefu unesibopho sokugcina idatha yevidiyo yakho futhi angalungiselelwa ukuthumela izexwayiso, izaziso, noma ngisho nokuqopha amavidiyo uma kutholwa imicimbi enyakazayo.

Umgomo wokugcinwa kwamafu wenze kwaba lula ukwakha uhlelo lokugada ividiyo ngezinjongo zokuthengisa. Manje sekungenzeka ukuthi ugcine inani elingenamkhawulo lezithombe ngaphandle kokudinga ihadiwe eyengeziwe noma ukukhathazeka ngokuphelelwa yisikhala somzimba.
 

Ukufinyelela kude

Esikhathini esedlule, wawuvame ukudinga ukufinyelela ngokomzimba ohlelweni lokuvikela. Ngokuxhuma amasistimu akho e-CCTV emafini, abasebenzisi abagunyaziwe bangafinyelela futhi babelane ngezithombe nganoma yisiphi isikhathi noma kuphi. Inzuzo enkulu yalolu hlobo lwesistimu ukuthi inika ibhizinisi lakho ukufinyelela kukho konke okurekhodiwe 24/7 noma yikuphi - ngisho noma ungekho ehhovisi!
 

Ukugcinwa kulula futhi kuyabiza

Ngaphezu kwalokho, izinsiza zokubhekwa kwevidiyo yamafu njengokugcinwa nokusatshalaliswa kokurekhodiwe zibuyekezwa ngokuzenzakalelayo, ngaphandle kokuzibandakanya komsebenzisi, okulula kakhulu kubasebenzisi. Isitoreji sevidiyo samafu kulula ukusimisa; ayidingi i-hardware noma i-IT kanye nochwepheshe bezokuphepha ukuze bagcine isistimu isebenza futhi.

 

• Izinzuzo Zohlelo Lokuvikela Olususelwe Efu Lokuhlanganisa Isixazululo Nesifaki

 

Ukufakwa nengqalasizinda

Kokubili izindleko zomkhiqizo ophathekayo nezabasebenzi zokufaka isixazululo sokulawula ukufinyelela okususelwe ku-IP esisingathwa ifu zibiza kancane kakhulu. Asikho iseva ephathekayo noma iseva ebonakalayo edingekayo, okuholela ekongeni kwezindleko ze-$1,000 kuya ku-$30,000 kuye ngosayizi wesistimu.

Isifaki akudingeki sifake isofthiwe kuseva ebonakalayo, silungise iseva endaweni yekhasimende noma ukukhathazeka uma ingxenye entsha yehadiwe kanye nesistimu yokusebenza kuthobela izinqubomgomo ze-IT zekhasimende.

Ekulawulweni kokufinyelela kwamafu, ihadiwe yokulawula ukufinyelela ingafakwa futhi ikhombe ngokushesha efwini, ihlolwe, futhi ilungiswe. Ngokusebenzisa isevisi yefu, ukufakwa kufushane, akuphazamisi kangako, futhi kudinga ingqalasizinda encane.

• Ukwehlisa izindleko zokulungisa okuqhubekayo

  Uma uhlelo lokulawula ukufinyelela selufakiwe, kunezindleko eziqhubekayo zokulunakekela. Lokhu kuhlanganisa ukuthuthukiswa kwesofthiwe namapeshi, ukuqinisekisa ukusebenza kahle kwehadiwe, futhi maduze. Ngesistimu yokulawula ukufinyelela okusekelwe emafini, cishe yonke le misebenzi yokulungisa ingenziwa kunoma iyiphi idivayisi noma kunini. Abahlinzeki beSoftware yokulawula ukufinyelela njengesevisi (i-SaaS) ngokuvamile bahlanganisa zonke izici zokuthuthukiswa nezibuyekezo zesofthiwe ezindlekweni zabo zonyaka zesofthiwe.
• 

Ngaphezu kwalokho, imininingwane yekhasimende ivamise ukusekelwa kumaseva aphathekayo amaningi kuyo yonke ingqalasizinda yefu, ngakho-ke asikho isidingo sokuthi isihlanganisi siye kusayithi, sinikeze izipele, sifake ukuthuthukiswa, bese silungiselela izibuyekezo ezifanele kumasevisi. Abahlanganisi abakhiphe amasistimu wamafu ngenxa yalokho babona inzuzo eyengeziwe, ukwaneliseka okukhulu kwamakhasimende, izindleko eziphansi ezingaphezulu, nokugcinwa kwamakhasimende okukhulu.
 

Ukuhlanganisa

Vula izixhumanisi zokuhlela izinhlelo zokusebenza (ama-API) anika amandla isilawuli sokufinyelela esihlanganisiwe kanye nesistimu yokungena ukuze ihlanganise nevidiyo, amakheshi, nezinye izinhlelo; amasistimu amaningi angahlanganiswa nokungena kunanini ngaphambili.

Noma yikuphi ukuhlanganiswa nobuchwepheshe bezinkampani zangaphandle kulula endaweni yesikhulumi esisekelwe efwini! Amasistimu avuliwe (esebenzisa ama-API) akwenza kube lula futhi kube lula ukuhlanganisa nezinhlelo nemikhiqizo yezinkampani zangaphandle, njengamathuluzi okuxhumana ebhizinisi avamile, njenge-CRM, i-ICT ne-ERP.

• Izinselelo Ezivamile Amabhizinisi Anamuhla abhekana nazo ekufakeni inkundla ye-Edge AI + Cloud Ekuvikelekeni Kokubhekwa Kwevidiyo

Ukuguquguquka okungalungile

Emkhakheni wokugadwa kwamavidiyo we-AI, ama-algorithms namadivayisi avame ukuba sesimweni sokuboshwa kakhulu. Kodwa ezinhlelweni zokusebenza, isistimu yokugada ividiyo idinga izinga elithile lokuguquguquka, okusho ukuthi ikhamera efanayo ivamise ukusetshenziswa ezimeni ezihlukene ezinama-algorithms ahlukene.

Ngamakhamera we-AI amaningi amanje, kunzima ukushintsha ama-algorithm uma eboshelwe ku-algorithm ethile. Ngakho-ke, izinkampani kufanele zichithe imali eningi emishinini emisha ukuxazulula izinkinga.

• Izinkinga zokunemba kwe-AI

  Ukuqaliswa kwe-AI ohlelweni lokubhekwa kwevidiyo kuthinteka kakhulu kukho kokubili ukubala kanye nezithombe. Ngenxa yemikhawulo yezingxenyekazi zekhompuyutha kanye nomthelela wemvelo yomhlaba wangempela, ukunemba kwesithombe sezinhlelo zokugada ze-AI ngokuvamile akuhambi kahle njengakulebhu. Kuzoba nomthelela omubi kulwazi lomsebenzisi kanye nokusetshenziswa kwangempela kwedatha.
  
  Amadivayisi aqondiwe onqenqemeni lwe-AI awavamile ukuba namandla noma asheshi ngokwanele ukuhlangabezana ngokugcwele nenkumbulo, ukusebenza, usayizi, nezidingo zokusetshenziswa kwamandla ze-Edge. Usayizi olinganiselwe kanye nomthamo wenkumbulo kuzophinde kuthinte ukukhethwa kwama-algorithms okufunda komshini.

• 

• Ukukhathazeka ngokuvikeleka kwedatha

  Indlela yokuhlinzeka ngezindlela zokuphepha ezanele ukuze kuvikelwe ulwazi lomsebenzisi futhi kuhlangatshezwane nezidingo zokuthobelana kuyinkinga eyinhloko okudingeka isistimu yokuvikela esekelwe emafini iyixazulule. Izingxenyekazi zekhompyutha ezithembekile ezinesofthiwe ethembekile zinhle, kodwa abantu abaningi bangase bakhathazeke ngokulahleka kwedatha noma ukudalulwa lapho itheminali ilayisha idatha emafini.

• 

• Isixazululo

Anviz IntelliSight Isixazululo singakwazi ukubona izinhlelo zokusebenza ezijwayelekile ze-AI ezisezingeni elingaphambili nge-11nm yakamuva ye-Qualcomm enamandla, i-2T yamandla ekhompiyutha ye-NPU. Ngasikhathi sinye, iyakwazi nokuqedela ngokushesha, uhlelo lokusebenza lwedatha lochwepheshe olusebenzayo ngenxa Anvizinkundla yesofthiwe esekwe emafini.

Le ndlela iyabiza futhi ilula, ngoba ayidingi noma yimiphi imishini eyengeziwe. Okuwukuphela kwezingxenyekazi zomzimba ezihilelekile Anviz amakhamera we-IP ahlakaniphile, ukurekhoda nokuthumela idatha efwini. Okurekhodiwe kwevidiyo kugcinwa kuseva ekude, engafinyelelwa nge-inthanethi.
 

Ukuguquguquka okuphezulu

The Anviz isixazululo sokubhekwa kwevidiyo - IntelliSight yamukela isofthiwe kanye nemodeli yokuhlukanisa ihadiwe, engabona ukushintshwa okuguquguqukayo kwama-algorithms ahlukahlukene we-AI. Anviz amatheminali afakwa ngaphambili ngamasethi ahlukahlukene ahlukene we-algorithm, futhi izinhlelo zokusebenza ezihlukene ze-algorithm zingenziwa zisebenze njengoba kudingeka. Kuthuthukisa kakhulu ukusebenza kahle kokuphatha kanye nesikhathi sokusebenzisa amakhamera we-AI futhi kunciphisa izindleko zokutshala izimali.
 

Ukunemba okuzinzile

I-algorithm ye-Neural network AI esekelwe ekuboneni isithombe ingathuthukisa ngempumelelo ikhono lokufunda elijulile kanye nokunemba kwe-algorithm. Anviz Ubuchwepheshe be-AI kumakhamera buhlanganisa ubuchwepheshe bokubona isithombe. Inquma okokuqala isimo esiguqukayo sesithombe, ilungise amapharamitha esithombe ukuze kuthuthukiswe ukuze kunikwe amandla ukubala kwe-AI, bese yenze ukuhlaziya i-AI. Ngakho-ke, impendulo yemiphumela yedatha ye-AI ihlale yenziwa ngaphansi kwesilinganiso esihlanganisiwe sesithombe, esithuthukisa kakhulu ukunemba kwe-AI.
 

Ukudluliswa kwedatha okuthembekile

Anviz Isixazululo samafu esithuthukisiwe sivikelekile ku-inthanethi ngokubethela kokuphela kuya ekupheleni, kusetshenziswa i-AES255 kanye ne-algorithm ye-HTTPS Encryption ukuvikela ukuphepha kwedatha njengoba i-Edge terminal ixhumana nefu. Ngaphezu kwalokho, yonke inqubo yokuxhumana ngamafu isekelwe ku- AnvizI-Control Protocol, ephinde ithuthukise ukusebenza kahle kokudluliswa kwedatha.