The invention relates to a method for training an artificial neural network, comprising the following steps: determining (S1) first data relating to a first road user (1) using a test vehicle sensor unit (11); determining (S2) second data relating to the first road user (1) using an external sensor unit (30); training (S3) the artificial neural network with the first data and with the second data as a respective input and with a predefined gauge for classifying normal or abnormal behaviour as a predefined output; and storing (S4) the trained neural network in a storage unit (3) for later execution in the vehicle (20), wherein the stored, trained, artificial neural network is designed to be executed with first data relating to a second road user (2) continuously detected by a vehicle sensor unit (21) of a vehicle (20) as the only input of the artificial neural network.
AUTOMATISIERTE ERKENNUNG EINES ANORMALEN VERHALTENS EINES VERKEHRSTEILNEHMERS
AUTOMATED DETECTION OF ABNORMAL BEHAVIOUR OF A ROAD USER
IDENTIFICATION AUTOMATISÉE D'UN COMPORTEMENT ANORMAL D'UN USAGER DE LA ROUTE
2022-07-06
Patent
Electronic Resource
German
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / B64C AEROPLANES , Flugzeuge / G05B Steuer- oder Regelsysteme allgemein , CONTROL OR REGULATING SYSTEMS IN GENERAL / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen |
AUTOMATISIERTE ERKENNUNG EINES ANORMALEN VERHALTENS EINES VERKEHRSTEILNEHMERS
European Patent Office | 2021
|Automatisierte Erkennung eines anormalen Verhaltens eines Verkehrsteilnehmers
European Patent Office | 2021
|VERFAHREN ZUM KLASSIFIZIEREN EINES VERHALTENS EINES VERKEHRSTEILNEHMERS
European Patent Office | 2020
|