Suna haɓaka software mai iya sanin abin da kuke tunani akai

hanyar yanar gizo

Ba tare da wata shakka ba, duniyar kimiyyar kwamfuta da koyon na'ura suna haɓaka da tsalle-tsalle. A wannan lokacin dole ne muyi magana game da sabon aikin da ƙungiyar masu bincike suka kasance na mashahuri Kamitani lab na Jami'ar Kyoto (Japan), waɗanda suka sami nasarar haɓaka software wanda, bisa ga gwaje-gwajen farko, yana iya karantawa da kuma sake haifuwa akan allo akan abin da mutum yake tunani.

Ba tare da wata shakka ba, muna magana ne game da ci gaban da ya fi dacewa fiye da yadda za mu iya gani tun da farko tun tasirin da aiki irin wannan zai iya yi a kowane yanki na rayuwarmu yana da girma ƙwarai Kuma duk wannan, ba shakka, ba tare da shiga cikin batutuwan soja ba, wani yanki wanda tabbas zai kasance mai sha'awar sanin yadda ake amfani da fasaha irin wannan.

neuron

Suna haɓaka cibiyar sadarwar ƙananan hanyoyi wanda zai iya fahimtar abin da kuke tunani akai

Kamar yadda ya gudana a cikin takardar da aka buga ta wata babbar mujalla mai sananniya kamar Science, a bayyane yake mafi rikitarwa kalubalen da ƙungiyar masu binciken Japan ta fuskanta ya kasance don cimmawa ƙirƙirar algorithm wanda ke iya fassarawa kuma, sama da duka, na sake maimaita hotunan da kwakwalwar mutum ta yi daidai bayan duba su akan allo. Hakanan, algorithm shima yana iya fassarawa da kuma maimaita abin da mutum ya tuna daga hotunan da suka gani a baya.

Akasin abin da yake iya zama alama, muna fuskantar mahimmin matakin da babu wanda ya sarrafa sakewa ta wata hanya. Don sanya wannan ci gaban a ɗan hangen nesa kuma mafi fahimtar abin da aka cimma, kawai gaya muku cewa ya zuwa yanzu, yunƙurin haɓaka algorithm mai iya yin wannan aikin ya kasance, don kiran su ko yaya, an iyakance. Bambanci tsakanin aikin da aka gudanar da kuma ma'anar da ƙungiyoyi daban-daban suka kai shine cewa wannan hanyar sadarwar ba kawai ta ƙara fahimta da kuma hayayyafar kwamfuta na hotunan kwakwalwa ba, amma kuma ya ba da izinin siffofin da kawai ke wanzuwa don fassara da kuma kwafa. tunanin mutum.

Masu aikin sa kai uku sun isa su horar da wannan software mai ban sha'awa

Kamar yadda aka bayyana, don taimakawa tare da horo da haɓaka cibiyar sadarwar, masu binciken da suka haɗu da ƙungiyar sun yanke shawarar amfani da su masu aikin sa kai guda uku masu hangen nesa ga waɗanda za su gabatar da hotuna daban-daban na ɗalibai kamar ɗabi'a, haruffa da sifofin geometric.

Manufar wannan gwajin ita ce lokacin kallon hotunan, za a samar da aiki a cikin kwakwalwar kwakwalwar kowane ɗayan waɗannan masu aikin sa kai wanda za a watsa da bincika su a cikin hanyar sadarwar ta jijiyoyin. Domin samun ci gaba da koya daga halayen daban-daban, masu aikin sa kai dole duba hotuna sama da 1.000 sau da yawa. Daga cikin hotunan, don mu sami kyakkyawan ra'ayi, zamu iya samun kifi, siffofi masu launuka masu sauƙi ko jirgin sama.

Hanyar hanyar sadarwa

Duk da irin ci gaban da kirkirar wannan manhaja ta kawo, masu kera ta har yanzu da sauran aiki a gabansu

Don yin rikodin kwakwalwar kowane ɗayan masu aikin sa kai, masu binciken sun yi amfani da fasahohi kamar su aiki maganadisu rawa, wanda ke auna jinin da yake gudana a wasu yankuna na kwakwalwa kuma don haka ya bayyana aikin jijiyoyin jiki. Tare da kowane ɗayan hotunan an bincika aikin kwakwalwar mutum. Godiya ga wannan babban aikin, daga karshe ya zama mai yiwuwa kwamfutar ta sami isasshen ƙarfin sake fasalin hoto daga aikin kwakwalwar da kowane mutum ke gabatarwa a kowane lokaci.

A matsayin cikakken bayani, kawai faɗi hakan sake gina hoto ba wani abu bane nan take ba amma cibiyar sadarwar ta bayyana abubuwa iri daya a kusan zagaye 200 tunda a zahiri fahimtar da mutum yake samu na abin da mutum ya yi game da hoton da aka gabatar ko aka tuna dole ne a gwada shi da waɗanda ya adana. A ƙarshe, azaman daki-daki, ba kawai hanyar sadarwar jijiyoyi ba ce ke iya yin kwatancen hoton kwakwalwa, amma yana samun ƙarin ƙwarewa ta hanyar keɓaɓɓiyar algorithm da aka aiwatar.


Abubuwan da ke cikin labarin suna bin ka'idodinmu na ka'idojin edita. Don yin rahoton kuskure danna a nan.

Kasance na farko don yin sharhi

Bar tsokaci

Your email address ba za a buga.

*

*

  1. Mai alhakin bayanan: Miguel Ángel Gatón
  2. Dalilin bayanan: Gudanar da SPAM, gudanar da sharhi.
  3. Halacci: Yarda da yarda
  4. Sadarwar bayanan: Ba za a sanar da wasu bayanan ga wasu kamfanoni ba sai ta hanyar wajibcin doka.
  5. Ajiye bayanai: Bayanin yanar gizo wanda Occentus Networks (EU) suka dauki nauyi
  6. Hakkoki: A kowane lokaci zaka iyakance, dawo da share bayanan ka.