In a continuous mobile communication scenario, e.g., between co-workers, participants may occasionally be located in the same space and thus hear each other naturally. To avoid hearing echoes, the audio transmission between these participants should be cut off. In this paper an acoustic co-location detection algorithm is proposed for the task, grouping participants together based solely on their microphone signals and mel-frequency cepstral coefficients thereof. The algorithm is tested both on recordings of different communication situations and in real time integrated into a voice-over-IP communication system. Tests on the recordings show that the algorithm works as intended, and the evaluation using the voice-over-IP conferencing system concludes that the algorithm improves the overall clarity of communication compared with not using the algorithm. The acoustic co-location detection algorithm thus proves a useful aid in continuous mobile communication systems.
(See document for exact affiliation information.)
https://aes2.org/publications/elibrary-page/?id=16795
(419KB)
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 Join the AES. If you need to check your member status, login to the Member Portal.
