The present invention relates to a computer-implemented method of training a convolutional neural network (CNN) and in particular a recurrent neural network (RNN) for autonomous driving, in particular of an autonomous car. Sets of training representations of surroundings of vehicles are recorded during real drives of real vehicles and corresponding driving states are recorded during the respective real drives of the real vehicles. Said recorded training representations of surroundings and corresponding driving states are used as sets of training data. This training data is used in training the CNN / RNN for autonomous driving of vehicles by providing driving states as output that can be used to derive control actions for the vehicles.
COMPUTER-IMPLEMENTED METHOD TRAINING A CNN / RNN FOR AUTONOMOUS DRIVING UTILIZING REINFORCEMENT LEARNING
COMPUTERIMPLEMENTIERTES VERFAHREN ZUM CNN/RNN-TRAINING FÜR AUTONOMES FAHREN UNTER VERWENDUNG VON VERSTÄRKUNGSLERNEN
PROCÉDÉ MIS EN UVRE PAR ORDINATEUR PERMETTANT D'ENTRAÎNER UN CNN / RNN POUR LA CONDUITE AUTONOME UTILISANT L'APPRENTISSAGE PAR RENFORCEMENT
2022-10-26
Patent
Electronic Resource
English
IPC: | G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion |
ArXiv | 2022
|Autonomous Driving with Deep Reinforcement Learning
SLUB | 2023
|AUTONOMOUS DRIVING REINFORCEMENT LEARNING APPARATUS AND METHOD OF VEHICLE
European Patent Office | 2019
|