arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2102.12394eess.AScs.SD
收藏

SEP-28 k:一个用于从口吃者的播客中检测口吃事件的数据集

SEP-28k: A Dataset for Stuttering Event Detection From Podcasts With People Who Stutter

Apple(苹果公司)

Colin Lea, Vikramjit Mitra, Aparna Joshi, Sachin Kajarekar, Jeffrey P. Bigham

摘要:

自动检测语音中的口吃事件的能力可以帮助语音病理学家随着时间的推移跟踪个人的流畅性,或者帮助改善具有非典型语音模式的人的语音识别系统。尽管人们对这一领域的兴趣越来越大,但现有的公共数据集太小,无法构建可推广的不流利检测系统,并且缺乏足够的注释。在这项工作中,我们介绍了播客中的口吃事件(SEP-28 k),这是一个包含超过28 k个片段的数据集,这些片段标记有五种事件类型,包括块,重复,声音重复,单词重复和感叹词。音频来自公共播客,主要由口吃的人采访其他口吃的人组成。我们在SEP-28 k和公共FluencyBank数据集上对一组声学模型进行了基准测试,并强调了简单地增加训练数据量如何将相对检测性能提高28%和24% F1。来自两个数据集的超过32,000个剪辑的注释将公开发布。

英文摘要:

The ability to automatically detect stuttering events in speech could help speech pathologists track an individual's fluency over time or help improve speech recognition systems for people with atypical speech patterns. Despite increasing interest in this area, existing public datasets are too small to build generalizable dysfluency detection systems and lack sufficient annotations. In this work, we introduce Stuttering Events in Podcasts (SEP-28k), a dataset containing over 28k clips labeled with five event types including blocks, prolongations, sound repetitions, word repetitions, and interjections. Audio comes from public podcasts largely consisting of people who stutter interviewing other people who stutter. We benchmark a set of acoustic models on SEP-28k and the public FluencyBank dataset and highlight how simply increasing the amount of training data improves relative detection performance by 28\% and 24\% F1 on each. Annotations from over 32k clips across both datasets will be publicly released.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑