德语合成语音与自然语音的差异性研究  

A Study of Differences between German Synthesized Speech and Natural Speech

作　　者：沙晓娅 SHA Xiaoya(School of Foreign Languages,East China University of Science and Technology,Shanghai,200237,China)

机构地区：[1]华东理工大学外国语学院,上海200237

出　　处：《文化创新比较研究》2024年第17期25-31,共7页Comparative Study of Cultural Innovation

摘　　要：随着人工智能技术的广泛应用,语音合成取得了重大技术进步。该文从语言学角度出发,采用语音处理领域客观定量的数据分析方法,对比分析了传统德语合成语音、新一代人工智能增强的合成语音及自然语音之间的差异及其原因,并进一步探讨了德语语音合成技术的改进方向。研究结果表明,相较于传统语音合成技术,基于人工智能的新一代德语语音合成技术在多项反映自然听感的指标上均展现出不错的提升。然而,由于德语本身独特的韵律特征,新一代人工智能合成语音在准确度和自然度方面与自然语音之间仍存在一定差距,具体体现在整体韵律、词间间隔以及音强变化等方面。With the wide application of artificial intelligence technology,speech synthesis has made significant technological progress.From a linguistic perspective,this paper adopts the method of objective quantitative data analysis in the field of speech processing to compare and analyse the differences and reasons between traditional German synthesized speech,new generation AI synthesized speech and natural speech,and further discusses the direction of improvement in German speech synthesis.The results of the study show that the new generation of AI-based German speech synthesis does show a good improvement over traditional German speech synthesis in terms of indicators reflecting the natural sense of hearing.However,due to the unique prosodic characteristics of the German language,there are still differences in accuracy and naturalness compared to natural speech in terms of overall rhythm,inter-word spacing and variations in sound intensity.

关 键 词：德语 合成语音 自然语音 人工智能 深度学习 TTS 

分 类 号：H33[语言文字—德语]