Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

LibriheavyMix: A 20,000-Hour Dataset for Single-Channel Reverberant Multi-Talker Speech Separation, ASR and Speaker Diarization

Zengrui Jin, Yifan Yang, Mohan Shi, Wei Kang, Xiaoyu Yang, Zengwei Yao, Fangjun Kuang, Liyong Guo, Lingwei Meng, Long Lin, Yong Xu, Shi-Xiong Zhang, Daniel Povey

TL;DR

This paper presents a large-scale far-field overlapping speech dataset, crafted to advance research in speech separation, recognition, and speaker diarization, and introduces a pipeline system encompassing speech separation, recognition, and diarization as a foundational benchmark.

Abstract

The evolving speech processing landscape is increasingly focused on complex scenarios like meetings or cocktail parties with multiple simultaneous speakers and far-field conditions. Existing methodologies for addressing these challenges fall into two categories: multi-channel and single-channel solutions. Single-channel approaches, notable for their generality and convenience, do not require specific information about microphone arrays. This paper presents a large-scale far-field overlapping speech dataset, crafted to advance research in speech separation, recognition, and speaker diarization. This dataset is a critical resource for decoding ``Who said What and When'' in multi-talker, reverberant environments, a daunting challenge in the field. Additionally, we introduce a pipeline system encompassing speech separation, recognition, and diarization as a foundational benchmark. Evaluations on the WHAMR! dataset validate the broad applicability of the proposed data.

LibriheavyMix: A 20,000-Hour Dataset for Single-Channel Reverberant Multi-Talker Speech Separation, ASR and Speaker Diarization

TL;DR

This paper presents a large-scale far-field overlapping speech dataset, crafted to advance research in speech separation, recognition, and speaker diarization, and introduces a pipeline system encompassing speech separation, recognition, and diarization as a foundational benchmark.

Abstract

The evolving speech processing landscape is increasingly focused on complex scenarios like meetings or cocktail parties with multiple simultaneous speakers and far-field conditions. Existing methodologies for addressing these challenges fall into two categories: multi-channel and single-channel solutions. Single-channel approaches, notable for their generality and convenience, do not require specific information about microphone arrays. This paper presents a large-scale far-field overlapping speech dataset, crafted to advance research in speech separation, recognition, and speaker diarization. This dataset is a critical resource for decoding ``Who said What and When'' in multi-talker, reverberant environments, a daunting challenge in the field. Additionally, we introduce a pipeline system encompassing speech separation, recognition, and diarization as a foundational benchmark. Evaluations on the WHAMR! dataset validate the broad applicability of the proposed data.
Paper Structure (12 sections, 5 equations, 1 figure, 6 tables, 2 algorithms)

This paper contains 12 sections, 5 equations, 1 figure, 6 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Data Simulation
  3. Baseline Systems
  4. Baseline for Multi-Talker ASR
  5. Baseline for Speech Separation
  6. Baseline for Speaker Diarization
  7. Experiments and Results
  8. Performance of Multi-Talker ASR Baseline
  9. Performance of Speech Separation Baseline
  10. Generalization on the WHAMR! dataset
  11. Performance of Speaker Diarization Baseline
  12. Conclusions

Figures (1)

  • Figure 1: Pipeline system of Separation, Diarization and ASR.