Deep learning and control algorithms of direct perception for autonomous driving

• Published in: Applied intelligence (Dordrecht, Netherlands) Vol. 51; no. 1; pp. 237 - 247
• Main Authors: Lee, Der-Hau, Chen, Kuan-Lin, Liou, Kuan-Han, Liu, Chang-Lun, Liu, Jinn-Liang
• Format: Journal Article
• Language: English
• Published: New York Springer US 2021; Springer Nature B.V
• Subjects: Algorithms; Artificial Intelligence; Artificial neural networks; Autonomous cars; Autonomous vehicles; Collisions; Computer Science; Computer simulation; Control algorithms; Deep learning; Dynamic stability; Evaluation; Indicators; Inference; Machine learning; Machines; Manufacturing; Mechanical Engineering; Perception; Processes; Race cars; Railroad cars; Regression models; Source code; Stability analysis; Traffic models; Traffic speed; Websites; Deep learning; Self-driving cars; Autonomous driving; Image perception; Control algorithms

We propose an end-to-end machine learning model that integrates multi-task (MT) learning, convolutional neural networks (CNNs), and control algorithms to achieve efficient inference and stable driving for self-driving cars. The CNN-MT model can simultaneously perform regression and classification ta...