In this paper an algorithm that improves side speaker attenuation for super directive beamformer like MVDR (Minimum Variance Distortionless Response) is presented. This technique can be utilized in a scenario where there are multiple people in a room intending to interact with an ASR- (automatic speech recognition) enabled device, e.g., smart speaker. The experiments show that the proposed solution gives a reduction of WER (word error rate) up to 23.93% calculated for command uttered by one user when a second user was treated as the side speaker.
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.