Publication:
KNOWLEDGE DISTILLATION USING GANS for FAST OBJECT DETECTION

Простая страница элемента
dc.contributor.authorGorbatsevich, V.
dc.contributor.authorMoiseenko, A.
dc.contributor.authorVizilter, Y.
dc.contributor.authorVygolov, O.
dc.contributor.authorFinogeev, E.
dc.date.accessioned2024-11-27T07:55:43Z
dc.date.available2024-11-27T07:55:43Z
dc.date.issued2020
dc.description.abstract© 2020 International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives.In this paper, we propose a new method for knowledge distilling based on generative adversarial networks. Discriminator CNNs is used as an adaptive knowledge distilling loss. In experiments, single shot multibox detector SSD based on MobileNet v2 and ShuffleNet v1 are used as student networks. Our tests showed AP and mAP improvement of more than 3% on PascalVOC and 1% on MS Coco datasets compared with the baseline algorithm without any architecture or dataset changes. The proposed approach is general and can be used not only with SSD but also with any type of object detection algorithms.
dc.format.extentС. 583-588
dc.identifier.citationKNOWLEDGE DISTILLATION USING GANS for FAST OBJECT DETECTION / Gorbatsevich, V. [et al.] // International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives. - 2020. - 43. - № B2. - P. 583-588. - 10.5194/isprs-archives-XLIII-B2-2020-583-2020
dc.identifier.doi10.5194/isprs-archives-XLIII-B2-2020-583-2020
dc.identifier.urihttps://www.doi.org/10.5194/isprs-archives-XLIII-B2-2020-583-2020
dc.identifier.urihttps://www.scopus.com/record/display.uri?eid=2-s2.0-85091065928&origin=resultslist
dc.identifier.urihttps://openrepository.mephi.ru/handle/123456789/22340
dc.relation.ispartofInternational Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives
dc.titleKNOWLEDGE DISTILLATION USING GANS for FAST OBJECT DETECTION
dc.typeConference Paper
dspace.entity.typePublication
oaire.citation.issueB2
oaire.citation.volume43
Файлы
Коллекции
Публикации