Coarse-to-fine Optimization for Speech Enhancement

被引:4
|
作者
Yao, Jian [1 ]
Al-Dahle, Ahmad [1 ]
机构
[1] Apple Inc, Cupertino, CA 95014 USA
来源
关键词
speech enhancement; coarse-to-fine; deep learning; generative model; discriminative model; dynamic perceptual loss;
D O I
10.21437/Interspeech.2019-2792
中图分类号
R36 [病理学]; R76 [耳鼻咽喉科学];
学科分类号
100104 ; 100213 ;
摘要
In this paper, we propose the coarse-to-fine optimization for the task of speech enhancement. Cosine similarity loss [1] has proven to be an effective metric to measure similarity of speech signals. However, due to the large variance of the enhanced speech with even the same cosine similarity loss in high dimensional space, a deep neural network learnt with this loss might not be able to predict enhanced speech with good quality. Our coarse-to-fine strategy optimizes the cosine similarity loss for different granularities so that more constraints are added to the prediction from high dimension to relatively low dimension. In this way, the enhanced speech will better resemble the clean speech. Experimental results show the effectiveness of our proposed coarse-to-fine optimization in both discriminative models and generative models. Moreover, we apply the coarse-to-fine strategy to the adversarial loss in generative adversarial network (GAN) and propose dynamic perceptual loss, which dynamically computes the adversarial loss from coarse resolution to fine resolution. Dynamic perceptual loss further improves the accuracy and achieves state-of-the-art results compared with other generative models.
引用
收藏
页码:2743 / 2747
页数:5
相关论文
共 50 条
  • [31] COARSE-TO-FINE VIDEO TEXT DETECTION
    Miao, Guangyi
    Huang, Qingming
    Jiang, Shuqiang
    Gao, Wen
    2008 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, VOLS 1-4, 2008, : 569 - +
  • [32] Coarse-to-Fine Region Selection and Matching
    Yang, Yanchao
    Lu, Zhaojin
    Sundaramoorthi, Ganesh
    2015 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2015, : 5051 - 5059
  • [33] Coarse-to-Fine Sparse Sequential Recommendation
    Li, Jiacheng
    Zhao, Tong
    Li, Jin
    Chan, Jim
    Faloutsos, Christos
    Karypis, George
    Pantel, Soo-Min
    McAuley, Julian
    PROCEEDINGS OF THE 45TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL (SIGIR '22), 2022, : 2082 - 2086
  • [34] Coarse-to-Fine Network for Crowd Counting
    Sun, Zhiyuan
    2022 IEEE INTERNATIONAL CONFERENCE ON ELECTRICAL ENGINEERING, BIG DATA AND ALGORITHMS (EEBDA), 2022, : 1342 - 1346
  • [35] Dataset condensation with coarse-to-fine regularization
    Jin, Hyundong
    Kim, Eunwoo
    PATTERN RECOGNITION LETTERS, 2025, 188 : 178 - 184
  • [36] Mesh optimization for surface approximation using an efficient coarse-to-fine evolutionary algorithm
    Huang, HL
    Ho, SY
    PROCEEDINGS OF THE 2001 CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1 AND 2, 2001, : 584 - 591
  • [37] A semi-supervised coarse-to-fine approach with bayesian optimization for lithology identification
    Xie, Yunxin
    Jin, Liangyu
    Zhu, Chenyang
    Wu, Siyu
    EARTH SCIENCE INFORMATICS, 2023, 16 (3) : 2285 - 2305
  • [38] A semi-supervised coarse-to-fine approach with bayesian optimization for lithology identification
    Yunxin Xie
    Liangyu Jin
    Chenyang Zhu
    Siyu Wu
    Earth Science Informatics, 2023, 16 : 2285 - 2305
  • [39] Mesh optimization for surface approximation using an efficient coarse-to-fine evolutionary algorithm
    Huang, HL
    Ho, SY
    PATTERN RECOGNITION, 2003, 36 (05) : 1065 - 1081
  • [40] Coarse-to-Fine Saliency Detection Based on Non-Subsampled Contourlet Transform Enhancement
    Liu Dongmei
    Chang Faliang
    ACTA OPTICA SINICA, 2019, 39 (01)