Naslov
Multi-Class U-Net for Segmentation of Non-biometric Identifiers

Autori
Hrkać, Tomislav ; Brkić, Karla ; Kalafatić, Zoran

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
IMVIP 2017 Irish Machine Vision and Image Processing Conference Proceedings / McDonald, John ; Markham, Charles ; Winslanley, Adam - Maynooth : Irish Pattern Recognition & Classification Society, 2017, 131-138

ISBN
978-0-9934207-2-6

Skup
IMVIP 2017 Irish Machine Vision and Image Processing

Mjesto i datum
Maynooth, Irska, 30.08.2017. - 01.09.2017

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
De-identification, Semantic segmentation, Deep learning

Sažetak
Ubiquity of image and video recording devices, as well as the increasing ease of sharing multimedia contents containing people without their permission induces serious privacy risks. Despite considerable efforts in research on de- identification of such contents, potentially identity-revealing information present in soft and non-biometric identifiers is often neglected. We propose an approach for segmentation of non- biometric identifiers intended for use in a de- identification pipeline that takes into account potentially identity-revealing characteristics such as dressing style, hairstyle, personal items, etc. The proposed approach is based on an adaptation of U-Net fully convolutional deep neural network architecture.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA