A Dataset for Provident Vehicle Detection at Night

1 code implementation • 27 May 2021 • Sascha Saralajew, Lars Ohnemus, Lukas Ewecker, Ebubekir Asan, Simon Isele, Stefan Roos

In this paper, we study the problem of how to map this intuitive human behavior to computer vision algorithms to detect oncoming vehicles at night just from the light reflections they cause by their headlights.

Object object-detection +1

Provident Vehicle Detection at Night: The PVDN Dataset

1 code implementation • 31 Dec 2020 • Lars Ohnemus, Lukas Ewecker, Ebubekir Asan, Stefan Roos, Simon Isele, Jakob Ketterer, Leopold Müller, Sascha Saralajew

As humans, we intuitively assume oncoming vehicles before the vehicles are actually physically visible by detecting light reflections caused by their headlamps.