JOINT DIRECTION AND PROXIMITY CLASSIFICATION OF OVERLAPPING SOUNDEVENTS FROM BINAURAL AUDIO

Sound source proximity and distance estimation are of great interest in many practical applications, since they provide significant information for acoustic scene analysis. As both tasks share complementary qualities, ensuring efficient interaction between these two is crucial for a complete picture of an aural environment. In this paper, we aim to investigate several ways of performing joint proximity and direction estimation from binaural recordings, both defined as coarse classification problems based on Deep Neural Networks (DNNs). Considering the limitations of binaural audio, we propose two methods of splitting the sphere into angular areas in order to obtain a set of directional classes. For each method we study different model types to acquire information about the direction-of-arrival (DoA). Finally, we propose various ways of combining the proximity and direction estimation problems into a joint task providing temporal information about the onsets and offsets of the appearing sources. Experiments are performed for a synthetic reverberant binaural dataset consisting of up to two overlapping sound events.

Simultaneous Target Classification and Moving Direction Estimation in Millimeter-Wave Radar System

In this study, we propose a method to identify the type of target and simultaneously determine its moving direction in a millimeter-wave radar system. First, using a frequency-modulated continuous wave (FMCW) radar sensor with the center frequency of 62 GHz, radar sensor data for a pedestrian, a cyclist, and a car are obtained in the test field. Then, a You Only Look Once (YOLO)-based network is trained with the sensor data to perform simultaneous target classification and moving direction estimation. To generate input data suitable for the deep learning-based classifier, a method of converting the radar detection result into an image form is also proposed. With the proposed method, we can identify the type of each target and its direction of movement with an accuracy of over 95%. Moreover, the pre-trained classifier shows an identification accuracy of 85% even for newly acquired data that have not been used for training.