Exploiting the directional coherence function for multichannel source extraction

doi:10.1016/j.specom.2021.01.002

	Exploiting the directional coherence function for multichannel source extraction
	Liang, Shan1 ; Li, Guanjun1,2 ; Nie, Shuai1 ; Yang, ZhanLei3 ; Liu, WenJu1 ; Tao, Jianhua1
发表期刊	SPEECH COMMUNICATION
ISSN	0167-6393
	2021-04-01
卷号	128 页码:1-14
通讯作者	Liu, WenJu(lwj@nlpr.ia.ac.cn)
摘要	The desired speech detector plays an important role for controlling the speech distortion in spatial filtering based speech enhancement algorithms. However, the conventional complex coherence(CC) based algorithms can only distinguish the coherent speech and diffuse noise. To improve the performance on the scenarios that both the coherent interference and diffuse noise are present, we propose a directional coherence function(DCF) based detector in this paper. Based on a pair of complementary filters which can suppress the diffuse noise and the coherent interference respectively, the DCF is computed as the normalized correlation between the filters? outputs. Meanwhile, the filters are solved by convex programming method and satisfy the constraints on speech distortionless and white noise gain(WNG). Consequently, the value of DCF will be close to 1 only for the desired speech dominated time-frequency(T-F) bins and much smaller than 1 for the noise or interference dominated T-F bins. To extract the desired speech, the DCF based Desired Speech Presence Probability(DSPP) is used to control the adaptation in general sidelobe canceler(GSC), and subsequently used as the post-filtering weight. Systematical experiments on several scenarios show that the proposed algorithm achieves significantly and consistently better noise suppression performance than the narrowband direction-of-arrival(DOA) estimates based algorithms.
关键词	Directional coherence function Coherent-to-Diffuse Ratio General sidelobe canceller Desired Speech Presence Probability
DOI	10.1016/j.specom.2021.01.002
收录类别	SCI
语种	英语
资助项目	National Key RD Plan of China[2016YFB1001404] ; China National Nature Science Foundation[61971419] ; China National Nature Science Foundation[61573357] ; China National Nature Science Foundation[61503382] ; China National Nature Science Foundation[61403370] ; China National Nature Science Foundation[61273267] ; China National Nature Science Foundation[91120303]
项目资助者	National Key RD Plan of China ; China National Nature Science Foundation
WOS研究方向	Acoustics ; Computer Science
WOS类目	Acoustics ; Computer Science, Interdisciplinary Applications
WOS记录号	WOS:000631974400001
出版者	ELSEVIER
七大方向——子方向分类	语音识别与合成
引用统计
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/44138
专题	多模态人工智能系统全国重点实验室_智能交互
通讯作者	Liu, WenJu
作者单位	1.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit NLPR, Beijing, Peoples R China 2.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing, Peoples R China 3.Huawei Technol, Shanghai, Peoples R China
第一作者单位	模式识别国家重点实验室
通讯作者单位	模式识别国家重点实验室
推荐引用方式 GB/T 7714	Liang, Shan,Li, Guanjun,Nie, Shuai,et al. Exploiting the directional coherence function for multichannel source extraction[J]. SPEECH COMMUNICATION,2021,128:1-14.
APA	Liang, Shan,Li, Guanjun,Nie, Shuai,Yang, ZhanLei,Liu, WenJu,&Tao, Jianhua.(2021).Exploiting the directional coherence function for multichannel source extraction.SPEECH COMMUNICATION,128,1-14.
MLA	Liang, Shan,et al."Exploiting the directional coherence function for multichannel source extraction".SPEECH COMMUNICATION 128(2021):1-14.