Using layer-wise training for Road Semantic Segmentation in Autonomous Cars

authored by: Shahrzad Shashaani, Mohammad Teshnehlab, Amirreza Khodadadian, Maryam Parvizi, Thomas Wick, Nima Noii
Abstract: A recently developed application of computer vision is pathfinding in self-driving cars. Semantic scene understanding and semantic segmentation, as subfields of computer vision, are widely used in autonomous driving. Semantic segmentation for pathfinding uses deep learning methods and various large sample datasets to train a proper model. Due to the importance of this task, accurate and robust models should be trained to perform properly in different lighting and weather conditions and in the presence of noisy input data. In this paper, we propose a novel learning method for semantic segmentation called layer-wise training and evaluate it on a light efficient structure called an efficient neural network (ENet). The results of the proposed learning method are compared with the classic learning approaches, including mIoU performance, network robustness to noise, and the possibility of reducing the size of the structure on two RGB image datasets on the road (CamVid) and off-road (Freiburg Forest) paths. Using this method partially eliminates the need for Transfer Learning. It also improves network performance when input is noisy.
Organisation(s): Institute of Applied Mathematics
Institute of Continuum Mechanics
External Organisation(s): K.N. Toosi University of Technology (KNTU)
Type: Article
Journal: IEEE ACCESS
Volume: 11
Pages: 46320 - 46329
No. of pages: 10
ISSN: 2169-3536
Publication date: 10.03.2023
Publication status: Published
Peer reviewed: Yes
ASJC Scopus subject areas: General Engineering, General Materials Science, Electrical and Electronic Engineering, General Computer Science
Electronic version(s): https://doi.org/10.1109/ACCESS.2023.3255988 (Access: Open)
: Details in the research portal "Research@Leibniz University"

BibTeX

@article{ffc7c85a44e14e9d90b86d13ee618ff7,
title = "Using layer-wise training for Road Semantic Segmentation in Autonomous Cars",
abstract = "A recently developed application of computer vision is pathfinding in self-driving cars. Semantic scene understanding and semantic segmentation, as subfields of computer vision, are widely used in autonomous driving. Semantic segmentation for pathfinding uses deep learning methods and various large sample datasets to train a proper model. Due to the importance of this task, accurate and robust models should be trained to perform properly in different lighting and weather conditions and in the presence of noisy input data. In this paper, we propose a novel learning method for semantic segmentation called layer-wise training and evaluate it on a light efficient structure called an efficient neural network (ENet). The results of the proposed learning method are compared with the classic learning approaches, including mIoU performance, network robustness to noise, and the possibility of reducing the size of the structure on two RGB image datasets on the road (CamVid) and off-road (Freiburg Forest) paths. Using this method partially eliminates the need for Transfer Learning. It also improves network performance when input is noisy.",
keywords = "Autonomous cars, Computer Vision, Convolution Neural Networks, Layer-wise trains, Semantic Segmentation, layer-wise trains, computer vision, convolution neural networks, semantic segmentation",
author = "Shahrzad Shashaani and Mohammad Teshnehlab and Amirreza Khodadadian and Maryam Parvizi and Thomas Wick and Nima Noii",
note = "Funding Information: The work of Maryam Parvizi was supported by the Alexander von Humboldt Foundation Project.",
year = "2023",
month = mar,
day = "10",
doi = "10.1109/ACCESS.2023.3255988",
language = "English",
volume = "11",
pages = "46320 -- 46329",
journal = "IEEE ACCESS",
issn = "2169-3536",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
}