Year: 2015
Type: dissertation
Abstract: In recent years, Deep Learning has emerged showing outstanding results for many different problems related to computer vision, machine learning and speech recognition. In this paper, we study the possibilities to apply convolutional neural networks (CNNs) and explore their power to address the pedestrian detection problem in the context of autonomous driving. We focus on creating a simple and robust framework based on the combination of a CNN architecture and a fast linear classifier. We show how the combination of these ingredients leads to a very accurate classifier, overcoming widespread techniques such as the HOG pedestrian detector and reaching state-of-the-art performance. Results from the wide range of experiments performed are analysed and compared on INRIA, one of the reference datasets for pedestrian detection.
Authors: Sergi Masip, Antonio M. López
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Funder: Nvidia
OpenAlex ID: W2280597037