?url_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rft.title=CLIPCleaner%3A+Cleaning+Noisy+Labels+with+CLIP&rft.creator=Feng%2C+Chen&rft.creator=Tzimiropoulos%2C+Georgios&rft.creator=Patras%2C+Ioannis&rft.description=Learning+with+Noisy+labels+(LNL)+poses+a+significant+challenge+for+the+Machine+Learning+community.+Some+of+the+most+widely+used+approaches+that+select+as+clean+samples+for+which+the+model+itself+(the+in-training+model)+has+high+confidence%2C+e.g.%2C+'small+loss'%2C+can+suffer+from+the+so+called+'self-confirmation'+bias.+This+bias+arises+because+the+in-training+model%2C+is+at+least+partially+trained+on+the+noisy+labels.+Furthermore%2C+in+the+classification+case%2C+an+additional+challenge+arises+because+some+of+the+label+noise+is+between+classes+that+are+visually+very+similar+(%60hard+noise').+This+paper+addresses+these+challenges+by+proposing+a+method+(CLIPCleaner)+that+leverages+CLIP%2C+a+powerful+Vision-Language+(VL)+model+for+constructing+a+zero-shot+classifier+for+efficient%2C+offline%2C+clean+sample+selection.+This+has+the+advantage+that+the+sample+selection+is+decoupled+from+the+in-training+model+and+that+the+sample+selection+is+aware+of+the+semantic+and+visual+similarities+between+the+classes+due+to+the+way+that+CLIP+is+trained.+We+provide+theoretical+justifications+and+empirical+evidence+to+demonstrate+the+advantages+of+CLIP+for+LNL+compared+to+conventional+pre-trained+models.+Compared+to+current+methods+that+combine+iterative+sample+selection+with+various+techniques%2C+CLIPCleaner+offers+a+simple%2C+single-step+approach+that+achieves+competitive+or+superior+performance+on+benchmark+datasets.+To+the+best+of+our+knowledge%2C+this+is+the+first+time+a+VL+model+has+been+used+for+sample+selection+to+address+the+problem+of+Learning+with+Noisy+Labels+(LNL)%2C+highlighting+their+potential+in+the+domain.&rft.subject=Sample+selection%2C+Noisy+Labels%2C+CLIP&rft.publisher=ACM&rft.date=2024-10-28&rft.type=Proceedings+paper&rft.language=eng&rft.source=+++++In%3A++Proceedings+of+the+32nd+ACM+International+Conference+on+Multimedia.++(pp.+pp.+876-885).++ACM+(2024)+++++&rft.format=text&rft.identifier=https%3A%2F%2Fdiscovery.ucl.ac.uk%2Fid%2Feprint%2F10199672%2F1%2FACMMM2024_CLIPCleaner_camera_ready.pdf&rft.identifier=https%3A%2F%2Fdiscovery.ucl.ac.uk%2Fid%2Feprint%2F10199672%2F&rft.rights=open