tudelft-iv / occluded_vehicle_acoustic_detection Goto Github PK

Is there any reason for the frequency-domain beamforming in the beamforming.py? Isn't it easier to just use beamforming in time domain so you don't have to "split" one wav-file in multiple HDF5-Files? Does acoular handle it differently? Or is there any other reason behind it?

If the detected vehicle approaches the intersection from the left, when the detected vehicle passes the intersection and enters the non line of sight area on the right again, what is the classification result of the model? That is, will a vehicle moving away from the right be recognized as a vehicle approaching from the right?

Is this sample log the only data sir or there is any unfiltered data related to it . I am working on this work in my dissertation for Masters if any more unfiltered data please share