Investigating Speaker Diarization of Endangered Language Data
Investigating Speaker Diarization of Endangered Language Data
复制标题
DOI：
--
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Gina-Anne Levow
中科院分区：
文献类型：
--
作者：
Gina-Anne Levow
发言人日记的任务旨在确定哪些发言人在录音中发言。这种功能可通过便利转录和半自动提取有用的元数据以丰富语言档案，帮助加快濒危语言的工作。然而，对于低资源或濒危语言的说话人日志化的工作很少。这项工作探讨了三种神经方法，以发言人日记适用于濒危语言档案中提取的数据集。我们发现最近的神经x向量模型与早期方法相比有一致的改进。我们还评估了影响模型和数据集性能的因素，重点关注濒危语言录音的挑战性特征。
The task of speaker diarization aims to determine which speakers spoke when in a recording. Such functionality could help to accelerate work in endangered languages by facilitating transcription and semi-automatically extracting useful meta-data to enrich language archives. However, there has been little work on speaker diarization for low-resource or endangered languages. This work explores three neural approaches to speaker diarization applied to data sets drawn from endangered language archives. We find consistent improvements for recent neural x-vector models over earlier approaches. We also assess the factors which impact performance across models and data sets, with a focus on the challenging characteristics of endangered language recordings.