Acoustic signal processing is very useful for autonomous rescue drones in disaster situations. Sound information can be used regardless of day or night, and continuous detection is possible for a long time with a small amount of resources. Through this code, a deep learning-based model is trained to detect the direction of a target source. This code is written in Python, based on Pytorch.
This opensource is a collaboration between Sogang University and NCSOFT. Additional information about the code can be found at the URL below.