Neural road semantic segmentation in driving scenarios

The purpose of this paper is to review the best semantic segmentation algorithms which involve deep neural networks and to compare one of the most used ones on a specific task for autonomous driving - road segmentation, regarding the accuracy and the time. We analyze the state of the art, then we propose a new dataset for road semantic segmentation consisting of images from our university campus, annotated by our team and taking into account the time of the day. We analyze the results of the road segmentation task obtained from the best segmentation models and we describe some metrics (accuracy, IoU) regarding the time of the day (day, dusk, night).