Sequence-level Speaker Change Detection with Difference-based Continuous Integrate-and-fire

TL;DR

This paper introduces a sequence transduction framework with a novel difference-based integrate-and-fire mechanism for speaker change detection, outperforming frame-level baselines on AMI and DIHARD-I datasets.

Contribution

It presents a new encoder-decoder approach with a difference-based integrate-and-fire mechanism for sequence-level speaker change detection, using weaker supervision.

Findings

Outperforms frame-level baseline methods.

Effective on AMI and DIHARD-I corpora.

Supports sequence-level supervision for speaker change detection.

Abstract

Speaker change detection is an important task in multi-party interactions such as meetings and conversations. In this paper, we address the speaker change detection task from the perspective of sequence transduction. Specifically, we propose a novel encoder-decoder framework that directly converts the input feature sequence to the speaker identity sequence. The difference-based continuous integrate-and-fire mechanism is designed to support this framework. It detects speaker changes by integrating the speaker difference between the encoder outputs frame-by-frame and transfers encoder outputs to segment-level speaker embeddings according to the detected speaker changes. The whole framework is supervised by the speaker identity sequence, a weaker label than the precise speaker change points. The experiments on the AMI and DIHARD-I corpora show that our sequence-level method consistently outperforms a strong frame-level baseline that uses the precise speaker change labels.