Computer implemented method and computer program for machine learning of the robustness of an acoustic classifier, acoustic classification system for automated driving systems and automated driving system

摘要

Computer-implemented method for machine learning of the robustness of an acoustic classifier (AK),wherein a driving system is controlled automatically depending on classifications and/or localizations of the acoustic classifier (AK), the procedure comprising the steps of providing initial input signals by means of a driving system acoustic sensor for the acoustic classifier (AK) (V1),Receiving interference (S) depending on the first input signals for deception detection, avoidance and/or protection and/or for improving the recognition and/or classification performance of the acoustic classifier (AK),where audibility of the interference is reduced (V2),Receiving second input data from an addition of the first input data and faults (V3),Entering combinations from the first and second input data into the acoustic classifier (AK) (V4) and machine learning of the combinations (V5),where the acoustic classifier (AK) learns to classify and/or locate acoustic events while becoming robust to interference.