Pedestrian and Vehicle detection in automotive embedded systems using deep neural networks

Pedestrian and vehicle detection by autonomous cars is an emerging area of research in the automotive community. The perception system of intelligent vehicles gathers data from sensing devices to understand and analyze traffic situations. This cognitive intelligence is required to make efficacious real-time decisions to avert imminent collisions with vulnerable traffic users such as humans, stranded or moving vehicles, cyclists or other static obstacles. This paper addresses the problem of people and vehicle detection using deep learning models such as convolutional neural networks. The results provided an incredible evidence that Deep neural networks have a significant potential for solving problems related to intelligent transportation systems. Higher rates of accuracy have been achieved by adding multiple hidden layers. The Keras based architecture along with tensorflow libraries have been used to implement the algorithm.