This engineering brief outlines how Machine Learning (ML) can be used to estimate objective sound source distance by examining both the temporal and spectral content of binaural signals. A simple ML algorithm is presented that is capable of predicting source distance to within half a meter in a previously unseen environment. This algorithm is trained using a selection of features extracted from synthesized binaural speech. This enables us to determine which of a selection of cues can be best used to predict sound source distance in binaural audio. The research presented can be seen not only as an exercise in ML but also as a means of investigating how binaural hearing works.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.
The Engineering Briefs at this Convention were selected on the basis of a submitted synopsis, ensuring that they are of interest to AES members, and are not overly commercial. These briefs have been reproduced from the authors' advance manuscripts, without editing, corrections, or consideration by the Review Board. The AES takes no responsibility for their contents. Paper copies are not available, but any member can freely access these briefs. Members are encouraged to provide comments that enhance their usefulness.