Gano Abun Kalmomi Buɗe
Gano abu na buɗe-ɓoye yana ba da damar samfurin ganowa da akwatin abubuwan da aka siffanta su ta hanyar rubutu na sabani, gami da nau'ikan da ba a taɓa gani ba a lokacin horo.
Dubawa
It matters because traditional detectors are locked to a fixed list of classes, while open-vocabulary models can detect almost anything you can name.
Zurfafa nutsewa
An horar da na'urori masu ganowa a kan rufaffiyar rufaffiyar rukunoni, sun ce azuzuwan 80 a cikin COCO, kuma ba za su iya gane wani 'abu' a wajen wannan jeri ba. Gane buɗewar ƙamus yana karya wanda ya iyakance ta hanyar daidaita fasalin yanki na gani tare da haɗin sararin hange-harshen hangen nesa, yawanci koya daga manyan nau'ikan rubutu-rubutu (kamar a cikin CLIP). Idan aka yi la'akari da cewa kuna samar da alamun rubutu, ƙirar tana haɗa waɗancan alamun, kuma ya dace da yankuna da aka gano zuwa kowane nau'in rubutun rubutu mafi kusa, don haka nau'ikan labari suna aiki muddin kuna iya siffanta su. Tsarin kamar ViLD, GLIP, OWL-ViT, Detic, da Grounding DINO sun haɓaka tsarin ta hanyar haɗa ƙasusuwan bayan ganowa tare da ƙasan harshe da kuma horo akan manyan bayanai masu rauni ko ƙasa.
Fahimtar Fasaha
Dabarar ita ce maye gurbin kafaffen Layer na rarrabawa tare da abubuwan rubutu. Maimakon koyon nau'in nau'in nau'in nau'in nau'in nau'in nau'in nau'in nau'in nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i nau'i) nau'i-nau'. rarrabuwa ya zama kwatankwacin kamanceceniya tsakanin fasalulluka na yanki da abubuwan da aka samar da sunaye ko jumlar nau'ikan da mai amfani ya samar. Saboda rukunin rubutun ya keɓanta ga kalmomin da ba a gani, musanya cikin sabbin igiyoyin lakabi a lokacin gwaji yana ba da damar gano nau'ikan da ba su nan daga bayanan horon akwatin.
Dabarun Tasiri
Gudu da sikelin
Kayayyakin AI na iya sarrafa aiki da bincike, ganowa, da ayyuka masu alama a sikelin.
Gina zaɓuɓɓuka
Ƙungiyoyin ƙirƙira za su iya samar da ra'ayoyi cikin sauri tare da ƙarancin bita da hannu.
Ƙungiya da aikin aiki
Ayyuka na iya amfani da siginar hoto da bidiyo waɗanda a baya suke da wahalar aiwatarwa.
Makomar Buɗe-Kamus Gane Abun Kalma
Gano ƙamus na buɗewa yana haɗuwa tare da ƙasa da rarrabuwa, inda jumlar sigar kyauta (ba kalmomi ɗaya kawai) ke gano abubuwa ba, kuma tare da tsarin gaggawa haɗe da ƙira kamar SAM don abin rufe fuska. Yi tsammanin daidaiton sifili mai ƙarfi, tsayi da ƙarin abubuwan rubutu na rubutu ('jariyar mug a bayan kwamfutar tafi-da-gidanka'), da madaidaicin haɗin gwiwa tare da mataimakan multimodal waɗanda ke gano kan buƙata. Yayin da horon rubutun-rubutu na sikelin yanar gizo ke inganta, layin tsakanin ganowa, dawo da, da fahimtar harshe zai ci gaba da yin dusashewa zuwa ƙasan gani na gaba ɗaya.
Aiwatar da Gaskiyar Duniya
Neman hotuna don ƙananan abubuwa ko na al'ada ta hanyar buga sunayensu ba tare da sake horarwa ba
Tsarin Robotics suna gano abu da mai amfani ya sanyawa suna a cikin yaren halitta kafin fahimtarsa
Saitin bayanai ta atomatik ta gano sabbin nau'ikan yawa daga jerin rubutu
Matsakaicin abun ciki wanda tutocin da aka siffanta abubuwan da ba su kasance a cikin alamun horo na asali
Hatsari & Tsare-tsare
Haƙƙoƙin hoto da yarda na iya zama haxarin doka idan ba a fayyace ba.
Ayyukan samfuri na iya bambanta a ko'ina cikin haske, ƙididdiga, da mahalli.
Ƙarya tabbataccen ƙila ba za a iya lura da shi ba sai dai idan an kula da ƙofofin amincewa.
Taswirar Hanya
Ƙayyade ma'auni na karɓa don daidaito, tunowa, da farashi na kuskure.
Gwada tare da bayanan da suka dace da ainihin yanayin samarwa.
Ƙara bita na ɗan adam don ƙarancin amincewa ko tsinkaya mai tasiri.
Bi diddigin ƙirar ƙira kuma sake ingantawa bayan canje-canjen kamara ko saitin bayanai.
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 Open-Vocabulary Object Detection 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
Gane Abu
Tambayoyin da ake yawan yi
What is Open-Vocabulary Object Detection?
Gano abu na buɗe-ɓoye yana ba da damar samfurin ganowa da akwatin abubuwan da aka siffanta su ta hanyar rubutu na sabani, gami da nau'ikan da ba a taɓa gani ba a lokacin horo. Yana da mahimmanci saboda an kulle masu gano na al'ada zuwa ƙayyadadden jerin azuzuwan, yayin da ƙirar ƙamus na iya gano kusan duk wani abu da za ku iya suna.
Menene ma'anar ikon gano abu buɗaɗɗen ƙamus?
Gano buɗaɗɗen ƙamus na iya gano nau'ikan da aka kayyade ta rubutu a lokacin gwaji, har ma waɗanda ba ta taɓa gani ba masu lakabi da akwatunan ɗaure.
Ta yaya waɗannan samfuran ke rarraba yankin da aka gano?
Suna tsara yankuna zuwa sararin hange-harshen hangen nesa kuma suna daidaita kowane yanki zuwa alamar rubutu mafi kusa.
Wanne kayan aikin riga-kafi ne ke ba da damar gano ƙamus?
Manyan bayanan rubutun hoto (kamar yadda tsarin CLIP ke amfani da shi) yana gina sararin haɗin gwiwa wanda zai ba da damar rubutu gabaɗaya zuwa sabbin nau'ikan.
Wane bangare ne aka maye gurbin idan aka kwatanta da rufaffiyar ganowa?
Instead of fixed class weight vectors, classification uses similarity to text embeddings, so new label strings can be added at test time.
Wanne ne a cikin waɗannan misalin buɗaɗɗen ƙamus ko ƙirar gano ƙasa?
Grounding DINO, tare da GLIP, OWL-ViT, ViLD, da Detic, sanannun ƙamus ne masu gano ƙamus ko ƙasa.