Evaluation of the robustness of deep features on the change detection problem
[ X ]
Tarih
2018
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Ieee
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
Deep Learning is a method which is employed for change detection as well as other image processing problems. Output extracted from various layers of the deep architecture can be employed to detect changes at different scales. In this study, output extracted from the layers of deep architecture is referred as deep features and the robustness of these features on the change detection problem are evaluated experimentally. As a result, it is observed that deep features, when used alone, could detect the change in images with steady background successfully but they were sensitive to dynamic background and camera jitter.
Açıklama
26th IEEE Signal Processing and Communications Applications Conference (SIU) -- MAY 02-05, 2018 -- Izmir, TURKEY
Anahtar Kelimeler
change detection, deep learning, deep features
Kaynak
2018 26Th Signal Processing And Communications Applications Conference (Siu)
WoS Q Değeri
N/A
Scopus Q Değeri
N/A
