作者: Alexandre Armand , David Filliat , Javier Ibanez-Guzman
DOI: 10.1109/ITSC.2013.6728466
关键词:
摘要: Each driver reacts differently to the same traffic conditions, however, most Advanced Driving Assistant Systems (ADAS) assume that all drivers are same. This paper proposes a method learn and model velocity profile follows as vehicle decelerates towards stop intersection. Gaussian Processes (GP), machine learning for non-linear regressions used profiles. It is shown GP well adapted such an application, using data recorded in real conditions. allow generation of normally distributed speed, given position on road. By comparison with generic profiles, benefits individual patterns ADAS issues presented.