Minimizing Supervision for Vision-Based Perception and Control in Autonomous Driving

Robinet, François

Reference : Minimizing Supervision for Vision-Based Perception and Control in Autonomous Driving


	Document type :	Dissertations and theses : Doctoral thesis
	Discipline(s) :	Engineering, computing & technology : Computer science
	Focus Areas :	Computational Sciences
	To cite this reference:	http://hdl.handle.net/10993/52331

Title :	Minimizing Supervision for Vision-Based Perception and Control in Autonomous Driving
Language :	English
Author, co-author :	Robinet, François [University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SEDAN >]
Publication date :	4-Oct-2022
Institution :	University of Luxembourg, Luxembourg, Luxembourg
Name of the degree :	Doctor of the University of Luxembourg in Computer Science
Number of pages :	130
Supervisor :	Frank, Raphaël
President of the jury :	State, Radu
Member of the jury :	Aouada, Djamila
	Hundt, Christian
	Müller, Christian
Keywords :	[en] autonomous driving ; self-supervision ; deep learning ; computer vision ; self-driving ; unsupervised learning
Abstract :	[en] The research presented in this dissertation focuses on reducing the need for supervision in two tasks related to autonomous driving: end-to-end steering and free space segmentation. For end-to-end steering, we devise a new regularization technique which relies on pixel-relevance heatmaps to force the steering model to focus on lane markings. This improves performance across a variety of offline metrics. In relation to this work, we publicly release the RoboBus dataset, which consists of extensive driving data recorded using a commercial bus on a cross-border public transport route on the Luxembourgish-French border. We also tackle pseudo-supervised free space segmentation from three different angles: (1) we propose a Stochastic Co-Teaching training scheme that explicitly attempts to filter out the noise in pseudo-labels, (2) we study the impact of self-training and of different data augmentation techniques, (3) we devise a novel pseudo-label generation method based on road plane distance estimation from approximate depth maps. Finally, we investigate semi-supervised free space estimation and find that combining our techniques with a restricted subset of labeled samples results in substantial improvements in IoU, Precision and Recall.
Research centres :	Interdisciplinary Centre for Security, Reliability and Trust (SnT) > Services and Data management research group (SEDAN)
Funders :	Fonds National de la Recherche - FnR
Target :	Researchers ; Professionals ; Students ; General public
Permalink :	http://hdl.handle.net/10993/52331
FnR project :	FnR ; FNR13301060 > François Robinet > MASSIVE > Machine Learning For Risk Assessment In Semi-autonomous Vehicles > 01/10/2018 > 31/08/2022 > 2018

File(s) associated to this reference

Fulltext file(s):

	File	Commentary	Version	Size	Access
Open access	thesis-francois-robinet.pdf	PhD Dissertation	Author postprint	8.04 MB	View/Open

All documents in ORBi^lu are protected by a user license.

University of Luxembourg Library | Feedback | Legal notices