complex-valued neural networks(CVNNs)have shown their excellent efficiency compared to their real counterparts in speech enhancement,image and signal *** throughout the years have made many efforts to improve the lear...
详细信息
complex-valued neural networks(CVNNs)have shown their excellent efficiency compared to their real counterparts in speech enhancement,image and signal *** throughout the years have made many efforts to improve the learningalgorithms and activation functions of *** CVNNs have proven to have better performance in handling the naturally complex-valued data and signals,this area of study will grow and expect the arrival of some effective improvements in the ***,there exists an obvious reason to provide a comprehensive survey paper that systematically collects and categorizes the advancement of *** this paper,we discuss and summarize the recent advances based on their learningalgorithms,activation functions,which is the most challenging part of building a CVNN,and ***,we outline the structure and applications of complex-valued convolutional,residual and recurrent neural ***,we also present some challenges and future research directions to facilitate the exploration of the ability of CVNNs.
To minimize speech distortion and residual noise, an optimal tradeoff between noise reduction and speech distortion needs to be considered. An optimal tradeoff method for single channel speech enhancement was presente...
详细信息
To minimize speech distortion and residual noise, an optimal tradeoff between noise reduction and speech distortion needs to be considered. An optimal tradeoff method for single channel speech enhancement was presented by solving a real-valued constrained optimization model in a recent literature. This paper proposes a new optimal tradeoff method for multichannel speech enhancement by solving a complex-valued optimization problem subject to a residual noise constraint with the masking threshold of the clean speech. An effective complex-valued multichannel learningalgorithm is developed and its convergence analysis is established completely in a complex domain. Experiment results confirm that the proposed multichannel speech enhancement algorithm outperforms several conventional algorithms in terms of both objective measures and subjective measures. (C) 2017 Elsevier B.V. All rights reserved.
暂无评论