arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2310.05352cs.CLcs.SDeess.AS
收藏

一瞥足矣:仅通过查看关键词提取目标句子

A Glance is Enough: Extract Target Sentence By Looking at A keyword

Harbin Institute of Technology(哈尔滨工业大学);Tsinghua University(清华大学);Beijing University of Posts and Telecommunications(北京邮电大学)

Ying Shi, Dong Wang, Lantian Li, Jiqing Han

AI总结:

该研究针对仅靠关键词从多人混合语音中提取目标语句的问题,提出基于Transformer架构结合交叉注意力机制的方法,在Librispeech上-3dB信噪比场景下PER达26%,远优于基线的96%。

摘要:

本文研究了仅以关键词作为输入,从多人交谈语音中提取目标句子的可能性。例如,在社会保障应用场景中,关键词可以是“help(救命)”,目标是识别出呼救者正在表述的内容,同时忽略其他说话者。为解决这一问题,我们提出采用Transformer架构对关键词和语音话语二者进行嵌入,随后借助交叉注意力机制,从拼接或重叠的语音中选取正确内容。在Librispeech数据集上的实验结果表明,我们提出的方法能够从信噪比为-3dB的高噪声混合语音中有效提取目标句子,其音素错误率(PER)达到26%,而基线系统的PER为96%。

英文摘要:

This paper investigates the possibility of extracting a target sentence from multi-talker speech using only a keyword as input. For example, in social security applications, the keyword might be "help", and the goal is to identify what the person who called for help is articulating while ignoring other speakers. To address this problem, we propose using the Transformer architecture to embed both the keyword and the speech utterance and then rely on the cross-attention mechanism to select the correct content from the concatenated or overlapping speech. Experimental results on Librispeech demonstrate that our proposed method can effectively extract target sentences from very noisy and mixed speech (SNR=-3dB), achieving a phone error rate (PER) of 26\%, compared to the baseline system's PER of 96%.

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。

↑