Conference paper

SOCHOR Jakub, ŠPAŇHEL Jakub, JURÁNEK Roman, DOBEŠ Petr and HEROUT Adam. Graph@FIT Submission to the NVIDIA AI City Challenge 2018. In: NVIDIA AI City Challenge 2018 (CVPRW). Salt Lake City: IEEE Computer Society, 2018, pp. 77-84. ISBN 978-1-5386-6100-0. Available from:
Publication language:english
Original title:Graph@FIT Submission to the NVIDIA AI City Challenge 2018
Title (cs):Graph@FIT příspěvek do NVIDIA AI City Challenge 2018
Proceedings:NVIDIA AI City Challenge 2018 (CVPRW)
Conference:IEEE/CVF Conference on Computer Vision and Pattern Recognition
Place:Salt Lake City, US
Publisher:IEEE Computer Society
vehicle speed measurement, vehicle re-identification, challenge, camera calibration
In our submission to the NVIDIA AI City Challenge, we address speed measurement of vehicles and vehicle re-identification. For both these tasks, we use a calibration method based on extracted vanishing points. We detect and track vehicles by a CNN-based detector and we construct 3D bounding boxes for all vehicles. For the speed measurement task, we estimate the speed from the movement of the bounding box in the 3D space using the calibration. Our approach to vehicle re-identification is based on extraction of visual features from "unpacked" images of the vehicles. The features are aggregated in temporal domain to obtain a single feature descriptor for the whole track. Furthermore, we utilize a validation network to improve the re-identification accuracy.
   author = {Jakub Sochor and Jakub {\v{S}}pa{\v{n}}hel and
	Roman Jur{\'{a}}nek and Petr Dobe{\v{s}} and Adam
   title = {Graph@FIT Submission to the NVIDIA AI City
	Challenge 2018},
   pages = {77--84},
   booktitle = {NVIDIA AI City Challenge 2018 (CVPRW)},
   year = 2018,
   location = {Salt Lake City, US},
   publisher = {IEEE Computer Society},
   ISBN = {978-1-5386-6100-0},
   doi = {10.1109/CVPRW.2018.00018},
   language = {english},
   url = {}

Your IPv4 address:
Switch to https