paint-brush
Uphando olutsha lutyhila ubuthathaka kwiZixhobo zoKhuseleko lwezobuGcisa ezidumileyo ngokuchasene nobusela be-AInge@escholar
163 ukufunda

Uphando olutsha lutyhila ubuthathaka kwiZixhobo zoKhuseleko lwezobuGcisa ezidumileyo ngokuchasene nobusela be-AI

Inde kakhulu; Ukufunda

Izixhobo zangoku zokukhusela i-AI ngokuchasene nokulinganisa isitayile azisebenzi. Iindlela ezilula zokulingisa zidlula ngokulula kuzo, zishiya amagcisa eveziwe. Amacebo amatsha okukhusela ayafuneka.
featured image - Uphando olutsha lutyhila ubuthathaka kwiZixhobo zoKhuseleko lwezobuGcisa ezidumileyo ngokuchasene nobusela be-AI
EScholar: Electronic Academic Papers for Scholars HackerNoon profile picture
0-item

Ababhali:

(1) Robert Honig, ETH Zurich ([email protected]);

(2) Javier Rando, ETH Zurich ([email protected]);

(3) uNicholas Carlini, i-Google DeepMind;

(4) UFlorian Tramer, ETH Zurich ([email protected]).

Itheyibhile yoQhagamshelwano

Isishwankathelo kunye 1. Intshayelelo

  1. Imvelaphi kunye nomsebenzi onxulumeneyo

  2. Imodeli yosongelo

  3. ISimbo Esomeleleyo Sokulingisa

  4. Ukuseta uMfuniselo

  5. Iziphumo

    6.1 Iziphumo eziphambili: Lonke uKhuseleko lujikeleziswa ngokulula

    6.2 Uhlalutyo

  6. Ingxoxo kunye neMpembelelo eBanzi, Imibulelo, kunye neeReferensi

A. Imizekelo yoBugcisa obuneenkcukacha

B. Robust Mimicry Izizukulwana

C. Iziphumo ezineenkcukacha

D. Umahluko ngeGlaze Finetuning

E. Iziphumo kwiGlaze 2.0

F. Iziphumo kwiNkungu v2

G. Iindlela zeSimbo sokuLingisa

H. IziKhuseli zeSimbo sokuLingisa esele zikhona

I. Iindlela zokulinganisa ezomeleleyo

J. Ukumisela uMfuniselo

K. Isifundo soMsebenzisi

L. Compute Resources

Abstract

Amagcisa axhalaba ngakumbi malunga nokuqhubela phambili kwimodeli yokuvelisa imifanekiso enokuthi iphindaphinde ngokusondeleyo iindlela zabo zobugcisa ezizodwa. Ukuphendula, izixhobo ezininzi zokukhusela ngokuchasene nesimbo sokulinganisa ziye zaphuhliswa ezibandakanya ukuphazamisana okuncinci kwimisebenzi yobugcisa epapashwe kwi-intanethi. Kulo msebenzi, sivavanya ukusebenza kokhuseleko oludumileyo-kunye nezigidi zokukhutshelwa-kwaye sibonisa ukuba zibonelela ngokhuseleko olungeyonyani. Sifumanisa ukuba umgudu ophantsi kunye neendlela "ezingekho kwi-shelf", ezifana nokunyuswa komfanekiso, zanele ukudala iindlela zokulinganisa ezinamandla ezithobisa kakhulu ukhuseleko olukhoyo. Ngophononongo lwabasebenzisi, sibonisa ukuba lonke ukhuselo olukhoyo lunokugqithwa ngokulula, lushiya amagcisa esesichengeni sokulinganisa isimbo. Siyalumkisa ukuba izixhobo ezisekelwe kwiingxaki ezichaseneyo azikwazi ukukhusela ngokuthembekileyo amagcisa ekusebenziseni kakubi i-AI yokuvelisa, kwaye sikhuthaza ukuphuhliswa kwezinye izisombululo zokukhusela.

1. Intshayelelo

Ukulinganisa isitayile sisicelo esidumileyo semifuziselo yesicatshulwa ukuya kumfanekiso. Kunikwe imifanekiso embalwa evela kumzobi, imodeli inokulungiswa ukuze kuveliswe imifanekiso emitsha ngolo hlobo (umzekelo, isiphekepheke kwisimbo sikaVan Gogh). Kodwa ukulinganisa isitayile kunokubangela ingozi enkulu ukuba kusetyenziswe kakubi. Ngokukodwa, amagcisa amaninzi anamhlanje anexhala lokuba abanye ngoku banokuvelisa imifanekiso ekopa isitayile sabo sobugcisa esahlukileyo, kwaye banokuba abathengi (Heikkila¨, 2022). Njengempendulo, ukhuseleko oluninzi luye lwaphuhliswa ukukhusela amagcisa kwisitayela sokulingisa (Shan et al., 2023a; Van Le et al., 2023; Liang et al., 2023). Olu khuselo longeza ukuphazamiseka okuchaseneyo kwimifanekiso epapashwa ngamagcisa kwi-intanethi, ukuze inyanzelise inkqubo yokulungiswa. Olu khuselo lufumene ingqwalasela ebalulekileyo kumajelo eendaba-kunye neempawu kwiNew York Times (Hill, 2023), CNN (Thorbecke, 2023) kunye neScientific American (Leffer, 2023)-kwaye zikhutshelwe ngamaxesha e-1M (Shan et al. , 2023a).


Nangona kunjalo, akukacaci ukuba ezi zixhobo ziwakhusela kangakanani amagcisa ngokuchasene nesimbo sokulingisa, ngakumbi ukuba umntu uzama ukubaphepha (Radiya-Dixit et al., 2021). Kulo msebenzi, sibonisa ukuba izixhobo zokukhusela isitayela se-state-of-the-art-Glaze (Shan et al., 2023a), Mist (Liang et al., 2023) kunye ne-Anti-DreamBooth (Van Le et al., 2023) -ayisebenzi xa ujongene neendlela ezilula zokulinganisa ezomeleleyo. Iindlela zokulinganisa ezinamandla esiziqwalaselayo zivela kwizicwangciso eziphantsi-ezifana nokusebenzisa iskripthi se-finetuning eyahlukileyo, okanye ukongeza ingxolo ye-Gaussian kwimifanekiso ngaphambi koqeqesho-kwizicwangciso zamanyathelo amaninzi adibanisa izixhobo ze-shelf. Siqinisekisa iziphumo zethu ngophando lomsebenzisi, olubonisa ukuba iindlela zokulinganisa ezinamandla zinokuvelisa iziphumo ezingabonakaliyo kumgangatho kwezo zifunyenwe kwimizobo engakhuselekanga (jonga uMzobo 1 womzekelo obonisa umzekeliso).


Sibonisa ukuba izixhobo ezikhoyo zokukhusela zibonelela ngombono wobuxoki wokhuseleko. Iindlela zethu zokulinganisa ezinamandla azifuni ukuphuhliswa kwezixhobo ezintsha okanye iindlela zokulungisa kakuhle, kodwa ngononophelo kuphela


Umzobo 1: Amagcisa asemngciphekweni wokulinganisa isimbo ukusuka kwiimodeli eziveliswayo zilungiswe kubugcisa babo. Izixhobo zokukhusela ezikhoyo zongeza ukuphazamiseka okuncinci kwimizobo epapashiweyo ukuthintela ukulingisa (uShan et al., 2023a; Liang et al., 2023; Van Le et al., 2023). Nangona kunjalo, ezi zikhuselo ziyasilela kwiindlela zokulinganisa ezomeleleyo, zinika imvakalelo yobuxoki yokhuseleko kwaye zishiya amagcisa esesichengeni. Umsebenzi wobugcisa ngu @nulevoy (Stas Voloshin), uphinde waphinda waphinda wavunyelwa.


ukudibanisa iindlela ezisemgangathweni zokusetyenzwa kwemifanekiso ezazisele zikho ngexesha ezi zixhobo zokhuseleko zaziqala ukwaziswa!. Ke ngoko, sikholelwa ukuba nabenzi bomgunyathi abanezakhono eziphantsi bebenokuziphepha ngokulula ezi zixhobo oko zasekwayo.


Nangona sivavanya izixhobo ezithile zokhuseleko ezikhoyo namhlanje, imida yokhuselo lokulinganisa isitayile. Amagcisa akwimeko engalunganga kuba kufuneka enze okuthile kuqala (oko kukuthi, xa umntu ekhuphele ubugcisa obukhuselweyo, ukhuselo alusenakuphinda lutshintshwe). Ukuze zisebenze, izixhobo ezikhuselayo zijongene nomsebenzi ocelomngeni wokudala izinto eziphazamisayo ezidlulisela kuyo nayiphi na indlela yokulungelelanisa, kwanabo bakhethwe ngokufanelekileyo kwixesha elizayo. Isigqibo esifanayo senziwe nguRadiya-Dixit et al. (Radiya-Dixit et al., 2021), owathi ukuxhatshazwa okuchaseneyo akukwazi ukukhusela abasebenzisi kwiinkqubo zokuqaphela ubuso. Ke ngoko siyalumkisa ukuba ubugcisa bokufunda ngomatshini ochaseneyo abuyi kukwazi ukukhusela amagcisa ngokuthembekileyo kukulinganisa isitayile sokuvelisa, kwaye sikhuthaze uphuhliso lwezinye iindlela zokukhusela amagcisa.


Sichaze iziphumo zethu kwizixhobo zokhuseleko ezichaphazelekayo phambi kokuba zipapashwe, ukuze zikwazi ukumisela eyona ndlela ilungileyo yokwenza kubasebenzisi abakhoyo.


Eli phepha li ifumaneka kwi-arxiv phantsi kweCC BY 4.0 ilayisenisi.


L O A D I N G
. . . comments & more!

About Author

EScholar: Electronic Academic Papers for Scholars HackerNoon profile picture
EScholar: Electronic Academic Papers for Scholars@escholar
We publish the best academic work (that's too often lost to peer reviews & the TA's desk) to the global tech community

ZIJONGE IIMPAWU

ELI NQAKU LINIKEZELWE KU...