arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1808.02504cs.CLeess.AS
收藏

设备导向的话语检测

Device-directed Utterance Detection

Amazon(亚马逊)

Sri Harish Mallidi, Roland Maas, Kyle Goehner, Ariya Rastrow, Spyros Matsoukas, Björn Hoffmeister

摘要:

在这项工作中,我们提出了一个分类器,用于区分设备导向的查询与语音助手的交互的背景下,从背景语音。应用程序包括拒绝错误唤醒或意外交互,以及启用唤醒词免费后续查询。考虑交互示例:$"Computer,~play~music", "Computer,~reduce~the~volume"$。在该交互中,用户需要重复用于第二查询的唤醒词($Computer$)。为了允许更自然的交互,设备可以在第一次查询之后立即重新进入监听状态(没有唤醒词重复),并接受或拒绝潜在的后续作为设备引导的或背景语音。该模型由两个长短期记忆(LSTM)神经网络组成,分别基于声学特征和自动语音识别(ASR)1-最佳假设进行训练。然后,训练前馈深度神经网络(DNN)以将从LSTM导出的声学和1-最佳嵌入与来自ASR解码器的特征相结合。实验结果表明,ASR解码器,声学嵌入和1-最佳嵌入产生的$9.3~\%$,$10.9~\%$和$20.1~\%$的等错误率(EER)分别。这些特征的组合导致$44~\%$相对改善,最终EER为$5.2~\%$。

英文摘要:

In this work, we propose a classifier for distinguishing device-directed queries from background speech in the context of interactions with voice assistants. Applications include rejection of false wake-ups or unintended interactions as well as enabling wake-word free follow-up queries. Consider the example interaction: $"Computer,~play~music", "Computer,~reduce~the~volume"$. In this interaction, the user needs to repeat the wake-word ($Computer$) for the second query. To allow for more natural interactions, the device could immediately re-enter listening state after the first query (without wake-word repetition) and accept or reject a potential follow-up as device-directed or background speech. The proposed model consists of two long short-term memory (LSTM) neural networks trained on acoustic features and automatic speech recognition (ASR) 1-best hypotheses, respectively. A feed-forward deep neural network (DNN) is then trained to combine the acoustic and 1-best embeddings, derived from the LSTMs, with features from the ASR decoder. Experimental results show that ASR decoder, acoustic embeddings, and 1-best embeddings yield an equal-error-rate (EER) of $9.3~\%$, $10.9~\%$ and $20.1~\%$, respectively. Combination of the features resulted in a $44~\%$ relative improvement and a final EER of $5.2~\%$.

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。

↑