机器人自身噪声环境下的自动语音识别

删除或更新信息，请邮件至freekaoyan#163.com(#换成@)

清华大学辅仁网/2017-07-07

机器人自身噪声环境下的自动语音识别

王建荣¹, 张句¹, 路文焕², 魏建国², 党建武¹

1. 天津大学计算机科学与技术学院, 天津 300072;
2. 天津大学软件学院, 天津 300072

Automatic speech recognition with robot noise

WANG Jianrong¹, ZHANG Ju¹, LU Wenhuan², WEI Jianguo², DANG Jianwu¹

1. School of Computer Science and Technology, Tianjin University, Tianjin 300072, China;
2. School of Computer Software, Tianjin University, Tianjin 300072, China

摘要:

输出: BibTeX | EndNote (RIS)

摘要当机器人移动身体任何部位时，都会不可避免地产生自身噪声。这些自身噪声由身体关节或其他硬件设备如风扇等引起。由于自身噪声距离机器人麦克风较近，较目标声源更容易被获取。该文根据机器人自身噪声种类，提出了一种将谱减法、关节噪声模板减法、基于标注区域的倒谱均值减法以及多条件训练相结合的方法，从而估计和抑制自身噪声。一系列实验证明了所提出的方法可以有效地减少自身噪声影响，提高语音识别的鲁棒性。

关键词 ：机器人,语音识别,语音增强

Abstract：Robots inevitably produce noise when they are moving any part of their body. Such noise is caused by the various body joint motors as well as the CPU cooling fans. Moreover, these noises are easily captured by the robots' microphones because they are closer to the microphones than the target speech source. This paper presents a de-noising method using the spectral subtraction, joint noise template substraction, labeled area cepstral mean substraction and multi-condition training to estimate and suppress robot noise. Tests show that this method significantly reduces the effect of robot noise which enhances the automatic speech recognition.

Key words：robot speech recognition speech enhancement

收稿日期: 2016-06-20 出版日期: 2017-02-21

ZTFLH:	TP242
	TN912.34

通讯作者:路文焕,副教授,E-mail:wenhuan@tju.edu.cnE-mail: wenhuan@tju.edu.cn

引用本文:

王建荣, 张句, 路文焕, 魏建国, 党建武. 机器人自身噪声环境下的自动语音识别[J]. 清华大学学报（自然科学版）, 2017, 57(2): 153-157.
WANG Jianrong, ZHANG Ju, LU Wenhuan, WEI Jianguo, DANG Jianwu. Automatic speech recognition with robot noise. Journal of Tsinghua University(Science and Technology), 2017, 57(2): 153-157.

链接本文:

http://jst.tsinghuajournals.com/CN/10.16511/j.cnki.qhdxxb.2017.22.007或 http://jst.tsinghuajournals.com/CN/Y2017/V57/I2/153

图表:

图1 机器人自身噪声环境下的语音样本

图2 处在不同幅度下的CPU散热风扇噪声

图3 AldebaranRobotics的NAO机器人

表1 不同技术以及技术组合方法在不同自身噪声下的语音识别结果(距离120cm)

图4 谱减法处理风扇噪声的前后对比

图5 关节噪声模板减法处理前后的对比

图6 基于标注区域倒谱均值减法与全局倒谱均值减法的对比

参考文献:

[1]	Ince G, Nakadai K, Rodemann T, et al. A hybrid framework for ego noise cancellation of a robot[C]//2010 IEEE International Conference on Robotics and Automation (ICRA). Piscataway, NJ:IEEE Press, 2010:3623-3628.
[2]	Breazeal C L. Designing Sociable Robots[M]. Boston, MA:MIT Press, 2004.
[3]	Miwa H, Okuchi T, Itoh K, et al. A new mental model for humanoid robots for human friendly communication introduction of learning system, mood vector and second order equations of emotion[C]//Proc 2003 IEEE International Conference on Robotics and Automation (ICRA). Piscataway, NJ:IEEE Press, 2003, 3:3588-3593.
[4]	Nakadai K, Lourens T, Okuno H G, et al. Active audition for humanoid[C]//Proc of the 17th National Conference on Artificial Intelligence and 12th Conference on Innovative Applications of Artificial Intelligence. Palo Alto, CA:AAAI Press, 2000:832-839.
[5]	Even J, Sawada H, Saruwatari H, et al. Semi-blind suppression of internal noise for hands-free robot spoken dialog system[C]//Proc 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Piscataway, NJ:IEEE Press, 2009:658-663.
[6]	Cohen I, Berdugo B. Speech enhancement for non-stationary noise environments[J]. Signal Processing, 2001, 81(1):2403-2418.
[7]	Cohen I, Berdugo B. Noise estimation by minima controlled recursive averaging for robust speech enhancement[J]. IEEE Signal Processing Letters, 2002, 9(1):12-15.
[8]	Ito A, Kanayama T, Suzuki M, et al. Internal noise suppression for speech recognition by small robots[J]. IEICE Technical Report Speech, 2005, 105:43-48.
[9]	Nishimura Y, Ishizuka M, Nakadai K, et al. Speech recognition for a humanoid with motor noise utilizing missing feature theory[C]//2006 6th IEEE-RAS International Conference on Humanoid Robots. Piscataway, NJ:IEEE Press, 2006:26-33.
[10]	Ince G, Nakadai K, Rodemann T, et al. Incremental learning for ego noise estimation of a robot[C]//2011 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Piscataway, NJ:IEEE Press, 2011:131-136.
[11]	Ince G, Nakadai K, Rodemann T, et al. Ego noise suppression of a robot using template subtraction[C]//Proceedings of the 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems. Piscataway, NJ:IEEE Press, 2009:199-204.
[12]	Boll S. Suppression of acoustic noise in speech using spectral subtraction[J]. Processing IEEE Transactions on Acoustics Speech & Signal, 1979, 27(2):113-120.
[13]	Viikki O, Bye D, Laurila K. A recursive feature vector normalization approach for robust speech recognition in noise[C]//Proc 1998 IEEE International Conference on Acoustics, Speech and Signal Processing. Piscataway, NJ:IEEE Press, 1998, 2:733-736.

[1]	甘振业, 陈浩, 杨鸿武. 结合EEMD与K-SVD字典训练的语音增强算法[J]. 清华大学学报（自然科学版）, 2017, 57(3): 286-292.
[2]	米吉提·阿不里米提, 艾克白尔·帕塔尔, 艾斯卡尔·艾木都拉. 基于层次化结构的语言模型单元集优化[J]. 清华大学学报（自然科学版）, 2017, 57(3): 257-263.
[3]	王国磊, 伊强, 缪东晶, 陈恳, 王力强. 面向机器人喷涂的多变量涂层厚度分布模型[J]. 清华大学学报（自然科学版）, 2017, 57(3): 324-330.
[4]	艾斯卡尔·肉孜, 殷实, 张之勇, 王东, 艾斯卡尔·艾木都拉, 郑方. THUYG-20:免费的维吾尔语语音数据库[J]. 清华大学学报（自然科学版）, 2017, 57(2): 182-187.
[5]	张鹏远, 计哲, 侯炜, 金鑫, 韩卫生. 小资源下语音识别算法设计与优化[J]. 清华大学学报（自然科学版）, 2017, 57(2): 147-152.
[6]	张继文, 刘莉, 陈恳. 基于AHRS反馈的仿人机器人步行稳定控制[J]. 清华大学学报（自然科学版）, 2016, 56(8): 818-823.
[7]	张继文, 刘莉, 陈恳. 小型仿人足球机器人MOS-7的系统设计及局部优化[J]. 清华大学学报（自然科学版）, 2016, 56(8): 811-817.
[8]	邢安昊, 张鹏远, 潘接林, 颜永红. 基于SVD的DNN裁剪方法和重训练[J]. 清华大学学报（自然科学版）, 2016, 56(7): 772-776.
[9]	罗磊, 陈恳, 杜峰坡, 马振书. 基于改进型粒子群算法的曲面匹配与位姿获取[J]. 清华大学学报（自然科学版）, 2015, 55(10): 1061-1066.
[10]	邵君奕,张传清,陈雁,陈恳. 用于空间内曲面喷涂的冗余度机器人轨迹规划方法[J]. 清华大学学报（自然科学版）, 2014, 54(6): 799-804.
[11]	潘玉龙, 王国磊, 朱丽, 陈雁, 陈恳. 管道喷涂机器人喷枪运动速度优化[J]. 清华大学学报（自然科学版）, 2014, 54(2): 212-216.
[12]	刘佳君, 孙振国, 张文增, 陈强. 两端吸附式爬壁机器人机械臂运动误差修正算法[J]. 清华大学学报（自然科学版）, 2014, 54(2): 185-190.