Pedestrian Detection Based on Light-Weighted Separable Convolution for Advanced Driver Assistance Systems
- PDF / 1,260,669 Bytes
- 14 Pages / 439.37 x 666.142 pts Page_size
- 97 Downloads / 229 Views
Pedestrian Detection Based on Light-Weighted Separable Convolution for Advanced Driver Assistance Systems Riadh Ayachi1
· Yahia Said1,2 · Abdessalem Ben Abdelaali1
Accepted: 4 October 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract The growth in the number of vehicles in the world makes it hard to safely share the environment with pedestrians. Pedestrian’s safety is an important task that needs to be granted in the traffic environment. New cars are equipped with advanced driver assistance systems (ADAS) with a variety of applications. Pedestrian detection application is one of the most important applications for an ADAS that needs to be enhanced. In this paper, we propose a pedestrian detection system to be implemented in an ADAS. The proposed system is based on convolutional neural networks thanks to its performance when solving computer vision applications. On the other side, the proposed system ensures real-time processing and high detection performance. The proposed system will be designed by tacking the advantage of building lightweight convolution blocks and model compression techniques to ensure an embedded implementation. Those blocks will guarantee high precision and fast processing speed. To train and evaluate the proposed system, we used the Caltech dataset. The evaluation of the proposed system resulted in 87% of mean average precision and an inference speed of 35 frames per second. Keywords Pedestrian detection · Advanced driver assistance systems · Convolutional neural networks · Lightweight convolution blocks
1 Introduction The revolution of technology has enabled many important applications for human safety. One of the most dangerous environments for humans is the traffic environment when pedestrians share the same area with vehicles. To ensure the safety of pedestrians, drivers must have a global overview of objects around the vehicle. Advanced Driver Assistance Systems (ADAS) are intelligent systems used in today’s cars to assist the driver and help him to control the
B
Riadh Ayachi [email protected]
1
Laboratory of Electronics and Microelectronics (EµE), Faculty of Sciences of Monastir, University of Monastir, Monastir, Tunisia
2
Electrical Engineering Department, College of Engineering, Northern Border University, Arar, Saudi Arabia
123
R. Ayachi et al.
car in dangerous situations. This intelligent system can take control of the vehicle when the driver does not present any reaction for a potential danger like a collision with pedestrians. So, the ADAS must contain a pedestrian detection system to warn the driver about the existence of a pedestrian in the road or to make a reaction by reducing the vehicle speed or totally stop it. Thus, the pedestrian detection system must detect pedestrians and make a reaction in real-time. A pedestrian detection system is based on processing images provided by cameras fixed outside of the vehicle. Recent advances in computer vision applications make it easier to develop accurate and fast image processing alg
Data Loading...
