The sequence is composed by two sets of images, named as the day and night sets, which refers to the moment of the day they were acquired. The first set contains 5990 frames and the second 5081, divided in training and testing sets each sequence.
When using this dataset in your research, we will be happy if you cite us
Yainuvis Socarras, Sebastian Ramos, David Vazquez, Antonio Lopez, Theo Gevers. Adapting Pedestrian Detection from Synthetic to Far Infrared Images. In ICCV – Workshop on Visual Domain Adaptation and Dataset Bias. Sydney, Australia, 2013
Elektra is an autonomous vehicle project developed by research groups from the CVC research center and the UAB and UPC universities.
Our main objectives are boosting the research in self-driving technology and forming highly qualified engineers.
We imagine cities with a centralized mobility system to move safely, comfortably, saving energy and minimizing congestion. The city controls a fleet of automated vehicles that cooperate among them and with the city infrastructure.
Our funding mainly comes from public competitive and technological transfer projects.
We would like to acknowledge to the Spanish project TRA2014-57088-C2-1-R and NVIDIA for the donated hardware.