Deep Residual-Dense Lattice Network for Speech Enhancement

被引:0
|
作者
Nikzad, Mohammad [1 ]
Nicolson, Aaron [1 ]
Gao, Yongsheng [1 ]
Zhou, Jun [1 ]
Paliwal, Kuldip K. [1 ]
Shang, Fanhua [2 ]
机构
[1] Griffith Univ, Inst Integrated & Intelligent Syst, Brisbane, Qld, Australia
[2] Xidian Univ, Sch Artificial Intelligence, Xian, Peoples R China
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Convolutional neural networks (CNNs) with residual links (ResNets) and causal dilated convolutional units have been the network of choice for deep learning approaches to speech enhancement. While residual links improve gradient flow during training, feature diminution of shallow layer outputs can occur due to repetitive summations with deeper layer outputs. One strategy to improve feature re-usage is to fuse both ResNets and densely connected CNNs (DenseNets). DenseNets, however, over-allocate parameters for feature re-usage. Motivated by this, we propose the residual-dense lattice network (RDL-Net), which is a new CNN for speech enhancement that employs both residual and dense aggregations without over-allocating parameters for feature re-usage. This is managed through the topology of the RDL blocks, which limit the number of outputs used for dense aggregations. Our extensive experimental investigation shows that RDL-Nets are able to achieve a higher speech enhancement performance than CNNs that employ residual and/or dense aggregations. RDL-Nets also use substantially fewer parameters and have a lower computational requirement. Furthermore, we demonstrate that RDL-Nets outperform many state-of-the-art deep learning approaches to speech enhancement.
引用
收藏
页码:8552 / 8559
页数:8
相关论文
共 50 条
  • [1] Deep Residual-Dense Attention Network for Image Super-Resolution
    Qin, Ding
    Gu, Xiaodong
    [J]. NEURAL INFORMATION PROCESSING, ICONIP 2019, PT V, 2019, 1143 : 3 - 10
  • [2] Deep residual-dense network based on bidirectional recurrent neural network for atrial fibrillation detection
    Laghari, Asif Ali
    Sun, Yanqiu
    Alhussein, Musaed
    Aurangzeb, Khursheed
    Anwar, Muhammad Shahid
    Rashid, Mamoon
    [J]. SCIENTIFIC REPORTS, 2023, 13 (01)
  • [3] Deep residual-dense network based on bidirectional recurrent neural network for atrial fibrillation detection
    Asif Ali Laghari
    Yanqiu Sun
    Musaed Alhussein
    Khursheed Aurangzeb
    Muhammad Shahid Anwar
    Mamoon Rashid
    [J]. Scientific Reports, 13
  • [4] Speech Enhancement via Residual Dense Generative Adversarial Network
    Zhou, Lin
    Zhong, Qiuyue
    Wang, Tianyi
    Lu, Siyuan
    Hu, Hongmei
    [J]. COMPUTER SYSTEMS SCIENCE AND ENGINEERING, 2021, 38 (03): : 279 - 289
  • [5] Deep Supervised Residual Dense Network for Underwater Image Enhancement
    Han, Yanling
    Huang, Lihua
    Hong, Zhonghua
    Cao, Shouqi
    Zhang, Yun
    Wang, Jing
    [J]. SENSORS, 2021, 21 (09)
  • [6] Automatic echocardiographic anomalies interpretation using a stacked residual-dense network model
    Nurmaini, Siti
    Sapitri, Ade Iriani
    Tutuko, Bambang
    Rachmatullah, Muhammad Naufal
    Rini, Dian Palupi
    Darmawahyuni, Annisa
    Firdaus, Firdaus
    Mandala, Satria
    Nova, Ria
    Bernolian, Nuswil
    [J]. BMC BIOINFORMATICS, 2023, 24 (01)
  • [7] SHO based Deep Residual network and hierarchical speech features for speech enhancement
    Bhosle M.R.
    Narayaswamy N.K.
    [J]. International Journal of Speech Technology, 2023, 26 (2) : 355 - 370
  • [8] Automatic echocardiographic anomalies interpretation using a stacked residual-dense network model
    Siti Nurmaini
    Ade Iriani Sapitri
    Bambang Tutuko
    Muhammad Naufal Rachmatullah
    Dian Palupi Rini
    Annisa Darmawahyuni
    Firdaus Firdaus
    Satria Mandala
    Ria Nova
    Nuswil Bernolian
    [J]. BMC Bioinformatics, 24
  • [9] Speech Enhancement via Mask-Mapping Based Residual Dense Network
    Zhou, Lin
    Chen, Xijin
    Wu, Chaoyan
    Zhong, Qiuyue
    Cheng, Xu
    Tang, Yibin
    [J]. CMC-COMPUTERS MATERIALS & CONTINUA, 2023, 74 (01): : 1259 - 1277
  • [10] Attention-driven residual-dense network for no-reference image quality assessment
    Zhang, Yang
    Wang, Changzhong
    Lv, Xiang
    Song, Yingnan
    [J]. SIGNAL IMAGE AND VIDEO PROCESSING, 2024, 18 (SUPPL 1) : 537 - 551