Pregled bibliografske jedinice broj: 857781
Convolutional Neural Networks for Croatian Traffic Signs Recognition
Convolutional Neural Networks for Croatian Traffic Signs Recognition // Proceedings of the Croatian Computer Vision Workshop CCVW 2014 / Lončarić, Sven ; Subašić, Marko (ur.).
Zagreb: Sveučilište u Zagrebu, 2014. str. 15-20 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 857781 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Convolutional Neural Networks for Croatian Traffic Signs Recognition
Autori
Vukotić, Vedran ; Krapac, Josip ; Šegvić, Siniša.
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Proceedings of the Croatian Computer Vision Workshop CCVW 2014
/ Lončarić, Sven ; Subašić, Marko - Zagreb : Sveučilište u Zagrebu, 2014, 15-20
Skup
Croatian Computer Vision Workshop
Mjesto i datum
Zagreb, Hrvatska, 16.09.2014
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
Convolutional networks, traffic sign recognition
Sažetak
We present an approach to recognition of Croatian traffic signs based on convolutional neural networks (CNNs). A library for quick prototyping of CNNs is first developed. An architecture similar to LeNet-5 is then created and tested on the MNIST dataset of handwritten digits where comparable results were obtained. We analyze the FER-MASTIF TS2010 dataset and propose a CNN architecture for traffic sign recognition. The presented experiments confirm the feasibility of CNNs for the defined task and suggest improvements to be made in order to improve recognition of Croatian traffic signs.
Izvorni jezik
Engleski
Znanstvena područja
Računarstvo
POVEZANOST RADA
Ustanove:
Fakultet elektrotehnike i računarstva, Zagreb
