Matches in SemOpenAlex for { <https://semopenalex.org/work/W3043138801> ?p ?o ?g. }
- W3043138801 abstract "Novelty detection, i.e., identifying whether a given sample is drawn from outside the training distribution, is essential for reliable machine learning. To this end, there have been many attempts at learning a representation well-suited for novelty detection and designing a score based on such representation. In this paper, we propose a simple, yet effective method named contrasting shifted instances (CSI), inspired by the recent success on contrastive learning of visual representations. Specifically, in addition to contrasting a given sample with other instances as in conventional contrastive learning methods, our training scheme contrasts the sample with distributionally-shifted augmentations of itself. Based on this, we propose a new detection score that is specific to the proposed training scheme. Our experiments demonstrate the superiority of our method under various novelty detection scenarios, including unlabeled one-class, unlabeled multi-class and labeled multi-class settings, with various image benchmark datasets." @default.
- W3043138801 created "2020-07-23" @default.
- W3043138801 creator A5020674537 @default.
- W3043138801 creator A5035397475 @default.
- W3043138801 creator A5045815823 @default.
- W3043138801 creator A5050194076 @default.
- W3043138801 date "2020-07-16" @default.
- W3043138801 modified "2023-09-27" @default.
- W3043138801 title "CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances" @default.
- W3043138801 cites W114276366 @default.
- W3043138801 cites W12634471 @default.
- W3043138801 cites W1576445103 @default.
- W3043138801 cites W1636210188 @default.
- W3043138801 cites W1797268635 @default.
- W3043138801 cites W1836465849 @default.
- W3043138801 cites W1977295328 @default.
- W3043138801 cites W1996796871 @default.
- W3043138801 cites W2017745767 @default.
- W3043138801 cites W2047643928 @default.
- W3043138801 cites W2097117768 @default.
- W3043138801 cites W2103378897 @default.
- W3043138801 cites W2105497548 @default.
- W3043138801 cites W2115627867 @default.
- W3043138801 cites W2122646361 @default.
- W3043138801 cites W2127218421 @default.
- W3043138801 cites W2137130182 @default.
- W3043138801 cites W2150856297 @default.
- W3043138801 cites W2187089797 @default.
- W3043138801 cites W2194775991 @default.
- W3043138801 cites W2254249950 @default.
- W3043138801 cites W2335728318 @default.
- W3043138801 cites W2462906003 @default.
- W3043138801 cites W2518108298 @default.
- W3043138801 cites W2531327146 @default.
- W3043138801 cites W2599354622 @default.
- W3043138801 cites W2622263826 @default.
- W3043138801 cites W2732026016 @default.
- W3043138801 cites W2746314669 @default.
- W3043138801 cites W2757910899 @default.
- W3043138801 cites W2786088545 @default.
- W3043138801 cites W2786599352 @default.
- W3043138801 cites W2798302089 @default.
- W3043138801 cites W2798991696 @default.
- W3043138801 cites W2803697594 @default.
- W3043138801 cites W2842511635 @default.
- W3043138801 cites W2860343088 @default.
- W3043138801 cites W2867167548 @default.
- W3043138801 cites W2896687685 @default.
- W3043138801 cites W2904981516 @default.
- W3043138801 cites W2912237282 @default.
- W3043138801 cites W2913300775 @default.
- W3043138801 cites W2913939497 @default.
- W3043138801 cites W2944828972 @default.
- W3043138801 cites W2948571844 @default.
- W3043138801 cites W2949517790 @default.
- W3043138801 cites W2962742544 @default.
- W3043138801 cites W2962749812 @default.
- W3043138801 cites W2963049059 @default.
- W3043138801 cites W2963154279 @default.
- W3043138801 cites W2963344330 @default.
- W3043138801 cites W2963384319 @default.
- W3043138801 cites W2963693742 @default.
- W3043138801 cites W2963773039 @default.
- W3043138801 cites W2964212410 @default.
- W3043138801 cites W2970607325 @default.
- W3043138801 cites W2970946347 @default.
- W3043138801 cites W2980402966 @default.
- W3043138801 cites W2989929945 @default.
- W3043138801 cites W2995085126 @default.
- W3043138801 cites W2995458562 @default.
- W3043138801 cites W2995981555 @default.
- W3043138801 cites W2996156127 @default.
- W3043138801 cites W2996179709 @default.
- W3043138801 cites W3009561768 @default.
- W3043138801 cites W3034337759 @default.
- W3043138801 cites W3034695001 @default.
- W3043138801 cites W3034978746 @default.
- W3043138801 cites W3035524453 @default.
- W3043138801 cites W3041609598 @default.
- W3043138801 cites W3043462782 @default.
- W3043138801 cites W3082701951 @default.
- W3043138801 cites W3098628719 @default.
- W3043138801 cites W3100345210 @default.
- W3043138801 cites W3106428938 @default.
- W3043138801 cites W3115293622 @default.
- W3043138801 cites W3118608800 @default.
- W3043138801 cites W3120167236 @default.
- W3043138801 cites W3122232370 @default.
- W3043138801 hasPublicationYear "2020" @default.
- W3043138801 type Work @default.
- W3043138801 sameAs 3043138801 @default.
- W3043138801 citedByCount "19" @default.
- W3043138801 countsByYear W30431388012020 @default.
- W3043138801 countsByYear W30431388012021 @default.
- W3043138801 countsByYear W30431388012023 @default.
- W3043138801 crossrefType "posted-content" @default.
- W3043138801 hasAuthorship W3043138801A5020674537 @default.
- W3043138801 hasAuthorship W3043138801A5035397475 @default.
- W3043138801 hasAuthorship W3043138801A5045815823 @default.
- W3043138801 hasAuthorship W3043138801A5050194076 @default.