The Speed Submission to DIHARD II: Contributions & Lessons Learned

TL;DR

This paper details the Speed team's speaker diarization systems for DIHARD II, highlighting system components, lessons learned, and performance improvements over baselines in challenging real-world scenarios.

Contribution

The paper introduces a robust diarization system with multiple components and insights into effective approaches, outperforming challenge baselines.

Findings

System significantly outperformed baselines.

Component analysis revealed key factors affecting performance.

Lessons learned inform future diarization system design.

Abstract

This paper describes the speaker diarization systems developed for the Second DIHARD Speech Diarization Challenge (DIHARD II) by the Speed team. Besides describing the system, which considerably outperformed the challenge baselines, we also focus on the lessons learned from numerous approaches that we tried for single and multi-channel systems. We present several components of our diarization system, including categorization of domains, speech enhancement, speech activity detection, speaker embeddings, clustering methods, resegmentation, and system fusion. We analyze and discuss the effect of each such component on the overall diarization performance within the realistic settings of the challenge.