Isku xirka Neural Networks
Isku xirka Neural Networks (CNNs) ayaa ah qaab dhismeedka faraska shaqada si loo fahmo sawirada.
Dulmar
They learn visual patterns by sliding small filters across a picture, which is why they power everything from face unlock to medical scan analysis.
quusid qoto dheer
CNN-ku wuxuu farsameeyaa sawirka isagoo simbiraya xayndaabyo yaryar oo miisaanno ah, oo loo yaqaan filtarrada ama kernels, oo dhan pixels-yada. Shaandheyn kastaa waxay sawirtaa hal qaab, sida gees, barar midab, ama gees. Lakabyada hore waxay ogaadaan sifooyinka fudud; lakabyada qoto dheer waxay isku daraan indhaha, taayirrada, ama qoraalka. Sababtoo ah isla shaandhaynta ayaa dib loo istcimaalayaa boos kasta (miisaanka wadaagga), CNN-ku wuxuu u baahan yahay cabbirro aad uga yar marka loo eego shabakad si buuxda ugu xidhan oo waxay arki kartaa bisad haddii ay ka muuqato bidix-sare ama midigta hoose. Lakabyada barkadaha ayaa hoos u dhigaya sawirka u dhexeeya tillaabooyinka, taasoo ka dhigaysa shabakada mid dhakhso badan oo dulqaad leh isbeddellada yaryar. Naqshadaynta calaamadaha sida LeNet, AlexNet (2012) , iyo ResNet waxay wadeen kobcinta waxbarashada qoto dheer, iyada oo AlexNet's ImageNet uu ku guuleystay inuu kiciyo casriga casriga ah ee garoonka.
Aragtida Farsamada
Hawlgalka xudunta u ah waa convolution: filter (dheh 3x3 miisaanka) ayaa lagu dul dahaadhay balastii pixels, miisaan kasta waxaa lagu dhuftey pixels, natiijadiina waxaa lagu soo koobay hal lambar. Simbiriirixidda filtarku waxay soo saartaa khariidad sifada. Laba fikradood ayaa tan ka dhigaya mid waxtar leh: wadaaga miisaanka (hal shaandhayn ayaa dib loo isticmaalay meel kasta) iyo isku xidhka maxalliga ah (neuron kastaa wuxuu arkaa gobol yar oo keliya). Isku-darka isku-dhafka, kala-soocida aan tooska ahayn sida ReLU, iyo isku-dubbaridku waxay u oggolaanaysaa shabakaddu inay dhisto kala sareyn astaamo muuqaal ah oo aan la taaban karin.
Saamaynta Istiraatijiyadeed
Go'aamo cad
Waxay kaa caawinaysaa inaad kala saartid sheegashooyinka farsamada cad iyo luqadda suuq-geynta.
Qiimaha iyo miisaaniyada
Waxaad waydiin kartaa su'aalo fulineed oo wanaagsan ka hor inta aadan lacag ama waqti bixin.
Kooxda iyo socodka shaqada
Kooxaha fahamka la wadaago waxay sameeyaan wax soo saar, siyaasad, iyo go'aano waxbarasho oo wanaagsan.
Mustaqbalka Shabakadaha Neural-ka ee Convolutional
CNN-yadu waxay ahaanayaan kuwa ugu sarreeya aragtida-waqtiga-dhabta ah iyo kheyraadka- xaddidan, sida kamaradaha taleefoonnada iyo aragtida is-wadidda, sababtoo ah waa dhakhso iyo xog hufan. Transformers Vision hadda waxay ku xafiiltamaan ama ku garaacaan xog-ururin waaweyn, sidaa darteed goobta ayaa ku urursan naqshadaha isku-dhafka ah ee isku daraya waxtarka convolution iyo dareenka caalamiga ah. Filo CNN-yadu inay ku sii jiraan aaladaha guntan iyo cidhifyada, sawirka caafimaadka halka xogtu ku yar tahay, iyo sida soosaarayaasha sifada waxtarka leh ee quudiya nidaamyada hab-dhismeedka badan sannadaha soo socda.
Dhaqangelinta Adduunka-dhabta ah
Ogaanshaha burooyinka, jabka, iyo xanuunka macaanka ee raajada, CT scans, iyo sawirada isha
Awood u siinta aqoonsiga wejiga furitaanka taleefanka iyo ku dhejinta sawirka abka sida Google Sawirrada
Akhrinta calamadaha dariiqyada, calamadaha haadka, iyo dadka lugaynaya ee nidaamka garashada gaadhiga iskood u wada
Si otomaatig ah ugu calaacalida alaabada cilladaysan khadadka isku-xidhka warshadda iyadoo la adeegsanayo kormeerka kamarada
Khatarta & Dariiqyada Ilaalada
Kooxo kala duwan ayaa laga yaabaa inay isla erey u isticmaalaan si kala duwan, marka hore u qeex baaxadda.
Tilmaamaha ayaa u ekaan kara kuwo xooggan halka waxqabadka dhabta ah ee dunidu aanu sinnayn.
In la iska indho tiro tayada xogta iyo qorshayaasha qiimayntu waxay inta badan abuurtaa natiijooyin jilicsan.
Qorshe Hawleedka Dhaqangelinta
Ka bilow qeexidda luqadda cad ee natiijada aad u baahan tahay.
Dooro hal cabbir guusha iyo hal xaalad guuldarro ka hor tijaabada.
Ku orod duuliye yar oo wata xogta matale, ee ma aha bandhig muuqaal ah.
Dukumeenti meesha ay ka caawiso Shabakadaha Dareemaha ee Convolutional iyo meelaha hababka fudud ay ka fiican yihiin.
Sii wad Sahaminta
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 Convolutional Neural Networks 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
Hagaha xiga
Shabakadaha Neural Graph
Su'aalaha soo noqnoqda
What is Convolutional Neural Networks?
Isku xirka Neural Networks (CNNs) ayaa ah qaab dhismeedka faraska shaqada si loo fahmo sawirada. Waxay bartaan qaababka aragga iyagoo isku simbiraya filtarrada yaryar ee sawirka, waana sababta ay awood ugu yeeshaan wax kasta min fur furka wejiga ilaa falanqaynta iskaanka caafimaadka.
Waa maxay shaqada ugu weyn ee shaandhada (kernel) ee CNN?
Shaandheeyaha waa shabag yar oo miisaan ah oo ku dul simbiraya sawirka, oo dhaqaajiya marka uu helo qaabka uu bartay, sida gees ama muuqaal.
Muxuu CNN uga baahan yahay cabbirro aad uga yar marka loo eego shabakada si buuxda ugu xiran sawirrada?
Miisaanka wadaaga macnaheedu waa hal shaandheyn yar ayaa lagu dabaqay meel kasta oo sawirka ah, markaa shabakadu waxay dib u isticmaashaa miisaan isku mid ah halkii ay ka baran lahayd kuwa goonida ah goob kasta oo pixels ah.
Muxuu sida caadiga ah sameeyaa lakabka isku-darka?
Barkadeynta (sida max pooling) waxay hoos u dhigtaa tusaalaha khariidadda sifada, iyadoo yaraynaysa xisaabinta waxayna ka dhigaysaa shabakada mid xasaasi u ah sida saxda ah ee astaantu uga muuqato.
CNN qoto dheer, sidee ayay astaamuhu guud ahaan uga beddelaan lakabyada hore ilaa lakabyada dambe?
Lakabyada hore waxay ogaadaan sifooyinka hoose sida geesaha iyo midabada; lakabyada qoto dheer waxay kuwan isku daraan fikrado heer sare ah sida weji ama qaybo shay.
Dhacdodee ayaa si weyn loogu tixgaliyay in la laaday horumarka casriga ah ee aqoonta qoto dheer ee aragtida?
Guushii cajiibka ahayd ee AlexNet 2012 ImageNet iyadoo la adeegsanayo CNN qoto dheer oo ku saabsan GPU-yada waxay ku qancisay cilmi-baarayaasha in barashada qoto dheer ay ka fiicnaan karto hababkii hore, taasoo dhalinaysa kobaca degdega ah ee goobta.