ASoBO: Attentive Beamformer Selection for Distant Speaker Diarization in Meetings

TL;DR

This paper introduces ASoBO, a self-attention-based method for selecting beamformer outputs to improve distant speaker diarization in meetings, achieving high accuracy and explainability.

Contribution

It presents a novel self-attention algorithm for beamformer selection that enhances feature extraction for speaker diarization in multi-microphone recordings.

Findings

14.5% DER on AISHELL-4 dataset

Self-attention weights correlate with speaker locations

Improved distant VAD, OSD, and SD performance

Abstract

Speaker Diarization (SD) aims at grouping speech segments that belong to the same speaker. This task is required in many speech-processing applications, such as rich meeting transcription. In this context, distant microphone arrays usually capture the audio signal. Beamforming, i.e., spatial filtering, is a common practice to process multi-microphone audio data. However, it often requires an explicit localization of the active source to steer the filter. This paper proposes a self-attention-based algorithm to select the output of a bank of fixed spatial filters. This method serves as a feature extractor for joint Voice Activity (VAD) and Overlapped Speech Detection (OSD). The speaker diarization is then inferred from the detected segments. The approach shows convincing distant VAD, OSD, and SD performance, e.g. 14.5% DER on the AISHELL-4 dataset. The analysis of the self-attention weights demonstrates their explainability, as they correlate with the speaker's angular locations.