Autonomous Audio-Supported Learning of Visual Classifiers for Traffic Monitoring

In this paper, using acoustic detection and classification of vehicles, the proposed autonomous self-learning framework generates scene adaptive vehicle classifiers without the need to hand label any video data.