ID 原文 译文
833 进而提取交警指挥手势中骨架的相对长度及其与重力加速度的夹角作为空间上下文特征， Then the relative lengths of the gesture skeletons and the an-gles between each skeleton w. r. t. gravity are extracted as the spatial context features of the traffic police gesture.
834 并引入长短时记忆网络(Long Short Term Memory，LSTM)提取交警指挥手势的时序特征; Mean-while, long-term memory (LSTM)is introduced to extract the temporal features of traffic police gestures.
835 最后，设计了融合空间上下文和时序特征的交警指挥手势识别机(Chinese Traffic Police Gesture Recognizer，CTPGR)，创建了包含 8 种交警指挥手势、时长约 2 小时的交警指挥手势视频库对 CTPGR 进行训练验证，并通过实验将 CTPGR 与已有交警手势识别算法进行了对比分析。 Finally, the Chi-nese traffic police gesture recognizer (CTPGR)based on CPM and LSTM is designed, and a two-hour traffic police gesturevideo is recorded to train and verify the CTPGR.
836 实验证明 CTPGR 可以快速准确地识别交警指挥手势，系统对复杂背景和动态交警指挥手势具有较强的适应能力。 Experimental results show that the CTPGR is capable of recognizing traffic police gestures with high accuracy, and is fast enough for online gesture prediction.
837 电子喉语音存在基频单一、发声机械、辐射噪声大等多种缺陷，这严重影响了电子喉语音可懂度和自然度，特别是对汉语普通话之类的声调语言，问题尤其严重。 The Electro-Laryngeal (EL)speech has some drawbacks such as single fundamental frequency, mechani-cal sound and large radiation noise. The drawbacks affect the intelligibility and naturalness of the EL speech.
838 汉语普通话电子喉语音识别存在辅音混淆的问题并且识别结果没有声调， Especially, the tonal language such as Mandarin EL speech would be worse understanding.
839 因此本文在识别结果的基础之上设计了拼音拼写修正器和声调标注工具，再结合基于 Tacotron-2 的TTS 实现了电子喉语音向正常语音的转换。 In this paper, the spelling corrector for pinyin and the tone labelling tool are designed to solve the problems that Mandarin EL speech recognition has some errors in consonantsand the recognition result has no tone. The result is synthesized into the healthy speech by TTS based on Tacotron-2.
840 客观评价实验结果表明，拼音拼写修正器可以提高拼音准确率，声调标注在有上下文的语义环境中具有较高准确率。 The ob-jective evaluation results show that the accuracy of pinyin spelling corrector has been improved;the accuracy of tone label-ling under contextual environment is very high.
841 主观听力测试结果表明，本文所提方法在不同语言水平上提高了汉语普通话电子喉语音的可懂度和自然度。 The subjective results shows the proposed method can improve the intelligi-bility and naturalness of the EL speech a lot.
842 研究结果表明，本文设计的方法可以将不带声调的电子喉语音转换为正常语音，相比于传统语音转换方法具有更高的性能。 The results illustrate that the proposed method can convert the EL speech with-out tone into the healthy speech. And the proposed method performs better than the traditional method based on speech signalprocessing.