跨语言自动语音识别转移中的语言差异从阿拉伯语到塔什利语
Georgia Zellou1, Mohamed Lahrouchi2
1University of California, Davis, Davis, USA. gzellou@ucdavis.edu.
Scientific reports
|January 3, 2024
概括
跨语言自动语音识别 (ASR) 从阿拉伯语转移Tashlhiyt显示了性能差异. 基于语言的方法对于将ASR适应于具有稀有结构的低资源语言至关重要.
科学领域:
- 计算语言学 计算语言学
- 语音技术 语言技术
- 低资源语言处理 低资源语言处理
背景情况:
- 由于资源有限,Tashlhiyt在语音技术方面面临着挑战.
- 现有的自动语音识别 (ASR) 系统往往不支持像Tashlhiyt.com这样的低资源语言.
- 从一个相关的,更高资源语言 (阿拉伯语) 的跨语言转移被探索为一个潜在的解决方案.
研究的目的:
- 评估将阿拉伯语ASR系统重新用于Tashlhiyt的可行性.
- 为了确定Tashlhiyt的ASR转移绩效中的系统差异.
- 研究语言特征 (例如,无母音词) 和说话风格对ASR准确性的影响.
主要方法:
- 调查了跨语言的ASR从商业阿拉伯语系统转移到Tashlhiyt.
- 使用文字错误率 (WER) 和莱文施泰因距离分析了性能.
- 在不同的词形和说话风格 (清晰与随意的演讲) 中比较结果.
主要成果:
- 在Tashlhiyt.中观察到ASR转移业绩的系统差异.
- 与清晰的演讲相比,随意说话模式的表现明显更差.
- 在清晰的演讲中,Tashlhiyt的无母音单词表现低于有母音单词.
结论:
- 跨语言的ASR传输是可行的,但表现出系统的性能差异,受说话模式和语音战术的影响.
- 基于语言的适应策略对于在资源较少的语言中有效的ASR重用至关重要,特别是那些具有罕见语言结构的语言.
- 这些发现有助于理解ASR性能差异和声信号的人机映射.
更多相关视频
相关概念视频
Improving Translational Accuracy
2.6K
2.6K
Language and Cognition
348
Language serves as a bridge between ideas and communication, influencing how individuals perceive and interact with the world. Psychologists have long debated whether language shapes thought or vice versa. This discussion gained grip with Edward Sapir and Benjamin Lee Whorf in the 1940s, who proposed that language determines thought, a concept known as linguistic determinism. They suggested that the vocabulary and structure of a language influence how its speakers think and perceive reality.
348
Lateralization
337
Brain lateralization refers to the division of mental processes and functions between the two hemispheres of the brain, a phenomenon that optimizes neural efficiency and underpins complex abilities in humans. This specialization allows each hemisphere to perform tasks where it has a comparative advantage, facilitating more refined cognitive capabilities across different domains.
337


