A Stereo Wind-Noise Suppressor with Null Beamforming and Frequency-Domain Noise Averaging

被引:0
|
作者
Kato, Masanori [1 ,3 ]
Sugiyama, Akihiko [1 ,2 ]
Komatsu, Tatsuya [1 ]
机构
[1] NEC Corp Ltd, Data Sci Res Labs, Kawasaki, Kanagawa 2118666, Japan
[2] Tokyo Metropolitan Univ, Fac Syst Design, Hino, Tokyo 1910065, Japan
[3] NEC Corp Ltd, Radio Applicat Guidance & Electroopt Div, Kawasaki, Kanagawa, Japan
关键词
wind noise; noise suppressor; stereo; null beamforming; frequency-domain averaging; ENHANCEMENT;
D O I
10.1587/transfun.E101.A.1631
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
This paper proposes a stereo wind-noise suppressor with frequency-domain noise averaging. A directional gain for diffuse wind noise is estimated frame by frame using a null beamformer based on interchannel phase difference which blocks the target signal. The wind-noise gain estimate is commonly multiplied by the input noisy signal to generate channel dependent wind noise estimates in order to cope with interchannel wind-noise imbalance. Interchannel phase agreement by target signal dominance or incidentally equal wind-noise phase, which leads to underestimation, is offset by averaging channel dependent wind-noise estimates along frequency. Evaluation results show that the mean PESQ score by the proposed wind-noise suppressor reaches 2.1 which is 0.2 higher than that by the wind-noise suppressor without averaging and 0.3 higher than that by a conventional monaural-noise suppressor with a statistically significant difference.
引用
收藏
页码:1631 / 1637
页数:7
相关论文
共 50 条
  • [1] A Wind-Noise Suppressor with SNR Based Wind-Noise Detection and Speech-Wind Discrimination
    Kato, Masanori
    Sugiyama, Akihiko
    IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES, 2018, E101A (10): : 1638 - 1645
  • [2] A Wind-Noise Suppressor with Wind-Burst Detection Based on a Stationary Noise Estimate.
    Sugiyama, Akihiko
    Kato, Masanori
    Miyahara, Ryoji
    2015 IEEE INTERNATIONAL CONFERENCE ON CONSUMER ELECTRONICS (ICCE), 2015, : 24 - 25
  • [3] A WIND-NOISE SUPPRESSOR BASED ON WIND-ONSET DETECTION AND SPECTRAL GAIN MODIFICATION
    Kato, Masanori
    Sugiyama, Akihiko
    2014 14TH INTERNATIONAL WORKSHOP ON ACOUSTIC SIGNAL ENHANCEMENT (IWAENC), 2014, : 144 - 148
  • [4] Binaural Wind-Noise Tracking with Steering Preset
    Thaleiser, Stefan
    Enzner, Gerald
    2022 30TH EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO 2022), 2022, : 70 - 74
  • [5] A low-complexity noise suppressor with nonuniform subbands and a frequency-domain highpass filter
    Kato, Masanori
    Sugiyama, Akihiko
    2006 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING, VOLS 1-13, 2006, : 473 - 476
  • [6] Frequency-domain prediction of turbofan noise radiation
    Özyörük, Y
    Alpman, E
    Ahuja, V
    Long, LN
    JOURNAL OF SOUND AND VIBRATION, 2004, 270 (4-5) : 933 - 950
  • [7] Frequency-domain method for separating signal and noise
    王正明
    段晓君
    Science in China(Series E:Technological Sciences), 2000, (01) : 9 - 16
  • [8] SIMULATION OF THE PHASE NOISE OF OSCILLATORS IN THE FREQUENCY-DOMAIN
    ANZILL, W
    KARTNER, FX
    RUSSER, P
    AEU-ARCHIV FUR ELEKTRONIK UND UBERTRAGUNGSTECHNIK-INTERNATIONAL JOURNAL OF ELECTRONICS AND COMMUNICATIONS, 1994, 48 (01): : 45 - 50
  • [9] PREDICTION OF PROPFAN NOISE BY A FREQUENCY-DOMAIN SCHEME
    GOUNET, H
    LEWY, S
    JOURNAL OF AIRCRAFT, 1988, 25 (05): : 428 - 435
  • [10] Frequency-domain method for separating signal and noise
    Zhengming Wang
    Xiaojun Duan
    Science in China Series E: Technological Sciences, 2000, 43 : 9 - 16