scholarly journals A Multi-Resolution Approach to GAN-Based Speech Enhancement

2021 ◽  
Vol 11 (2) ◽  
pp. 721
Author(s):  
Hyung Yong Kim ◽  
Ji Won Yoon ◽  
Sung Jun Cheon ◽  
Woo Hyun Kang ◽  
Nam Soo Kim
Keyword(s):  
Speech Enhancement ◽  
Optimal Solution ◽  
Experimental Results ◽  
Generative Adversarial Networks ◽  
The Real ◽  
Multi Scale ◽  
Adversarial Networks ◽  
Speech Characteristics ◽  
Conventional Methods ◽  
Convex Property

Recently, generative adversarial networks (GANs) have been successfully applied to speech enhancement. However, there still remain two issues that need to be addressed: (1) GAN-based training is typically unstable due to its non-convex property, and (2) most of the conventional methods do not fully take advantage of the speech characteristics, which could result in a sub-optimal solution. In order to deal with these problems, we propose a progressive generator that can handle the speech in a multi-resolution fashion. Additionally, we propose a multi-scale discriminator that discriminates the real and generated speech at various sampling rates to stabilize GAN training. The proposed structure was compared with the conventional GAN-based speech enhancement algorithms using the VoiceBank-DEMAND dataset. Experimental results showed that the proposed approach can make the training faster and more stable, which improves the performance on various metrics for speech enhancement.

scholarly journals Fractional Wavelet-Based Generative Scattering Networks

2021 ◽  
Vol 15 ◽  
Author(s):  
Jiasong Wu ◽  
Xiang Qiu ◽  
Jing Zhang ◽  
Fuzhi Wu ◽  
Youyong Kong ◽  
...  
Keyword(s):  
Dimensionality Reduction ◽  
Reduction Method ◽  
Gaussian White Noise ◽  
Principal Component ◽  
Experimental Results ◽  
Generative Adversarial Networks ◽  
Image Generation ◽  
Adversarial Networks ◽  
Dimensionality Reduction Method

Generative adversarial networks and variational autoencoders (VAEs) provide impressive image generation from Gaussian white noise, but both are difficult to train, since they need a generator (or encoder) and a discriminator (or decoder) to be trained simultaneously, which can easily lead to unstable training. To solve or alleviate these synchronous training problems of generative adversarial networks (GANs) and VAEs, researchers recently proposed generative scattering networks (GSNs), which use wavelet scattering networks (ScatNets) as the encoder to obtain features (or ScatNet embeddings) and convolutional neural networks (CNNs) as the decoder to generate an image. The advantage of GSNs is that the parameters of ScatNets do not need to be learned, while the disadvantage of GSNs is that their ability to obtain representations of ScatNets is slightly weaker than that of CNNs. In addition, the dimensionality reduction method of principal component analysis (PCA) can easily lead to overfitting in the training of GSNs and, therefore, affect the quality of generated images in the testing process. To further improve the quality of generated images while keeping the advantages of GSNs, this study proposes generative fractional scattering networks (GFRSNs), which use more expressive fractional wavelet scattering networks (FrScatNets), instead of ScatNets as the encoder to obtain features (or FrScatNet embeddings) and use similar CNNs of GSNs as the decoder to generate an image. Additionally, this study develops a new dimensionality reduction method named feature-map fusion (FMF) instead of performing PCA to better retain the information of FrScatNets,; it also discusses the effect of image fusion on the quality of the generated image. The experimental results obtained on the CIFAR-10 and CelebA datasets show that the proposed GFRSNs can lead to better generated images than the original GSNs on testing datasets. The experimental results of the proposed GFRSNs with deep convolutional GAN (DCGAN), progressive GAN (PGAN), and CycleGAN are also given.

scholarly journals Parkinson’s Disease EMG Data Augmentation and Simulation with DCGANs and Style Transfer

Sensors ◽  
2020 ◽  
Vol 20 (9) ◽  
pp. 2605 ◽  
Author(s):  
Rafael Anicet Zanini ◽  
Esther Luna Colombini
Keyword(s):  
Parkinson’S Disease ◽  
Parkinson's Disease ◽  
Data Augmentation ◽  
Experimental Results ◽  
Generative Adversarial Networks ◽  
Style Transfer ◽  
Treatment Approaches ◽  
Adversarial Networks

This paper proposes two new data augmentation approaches based on Deep Convolutional Generative Adversarial Networks (DCGANs) and Style Transfer for augmenting Parkinson’s Disease (PD) electromyography (EMG) signals. The experimental results indicate that the proposed models can adapt to different frequencies and amplitudes of tremor, simulating each patient’s tremor patterns and extending them to different sets of movement protocols. Therefore, one could use these models for extending the existing patient dataset and generating tremor simulations for validating treatment approaches on different movement scenarios.

scholarly journals MW-ACGAN: Generating Multiscale High-Resolution SAR Images for Ship Detection

Sensors ◽  
2020 ◽  
Vol 20 (22) ◽  
pp. 6673
Author(s):  
Lichuan Zou ◽  
Hong Zhang ◽  
Chao Wang ◽  
Fan Wu ◽  
Feng Gu
Keyword(s):  
High Resolution ◽  
Small Sample ◽  
Generative Adversarial Networks ◽  
Detection Accuracy ◽  
Data Set ◽  
Ship Detection ◽  
Multi Scale ◽  
Adversarial Networks ◽  
Sar Data ◽  
Composite Data

In high-resolution Synthetic Aperture Radar (SAR) ship detection, the number of SAR samples seriously affects the performance of the algorithms based on deep learning. In this paper, aiming at the application requirements of high-resolution ship detection in small samples, a high-resolution SAR ship detection method combining an improved sample generation network, Multiscale Wasserstein Auxiliary Classifier Generative Adversarial Networks (MW-ACGAN) and the Yolo v3 network is proposed. Firstly, the multi-scale Wasserstein distance and gradient penalty loss are used to improve the original Auxiliary Classifier Generative Adversarial Networks (ACGAN), so that the improved network can stably generate high-resolution SAR ship images. Secondly, the multi-scale loss term is added to the network, so the multi-scale image output layers are added, and multi-scale SAR ship images can be generated. Then, the original ship data set and the generated data are combined into a composite data set to train the Yolo v3 target detection network, so as to solve the problem of low detection accuracy under small sample data set. The experimental results of Gaofen-3 (GF-3) 3 m SAR data show that the MW-ACGAN network can generate multi-scale and multi-class ship slices, and the confidence level of ResNet18 is higher than that of ACGAN network, with an average score of 0.91. The detection results of Yolo v3 network model show that the detection accuracy trained by the composite data set is as high as 94%, which is far better than that trained only by the original SAR data set. These results show that our method can make the best use of the original data set, improve the accuracy of ship detection.

scholarly journals Generative Adversarial Network for Image Super-Resolution Combining Texture Loss

2020 ◽  
Vol 10 (5) ◽  
pp. 1729 ◽  
Author(s):  
Yuning Jiang ◽  
Jinhua Li
Keyword(s):  
Loss Function ◽  
Network Structure ◽  
Similarity Index ◽  
Super Resolution ◽  
Experimental Results ◽  
Generative Adversarial Networks ◽  
Generative Adversarial Network ◽  
Visual Evaluation ◽  
Low Frequencies ◽  
Adversarial Networks

Objective: Super-resolution reconstruction is an increasingly important area in computer vision. To alleviate the problems that super-resolution reconstruction models based on generative adversarial networks are difficult to train and contain artifacts in reconstruction results, we propose a novel and improved algorithm. Methods: This paper presented TSRGAN (Super-Resolution Generative Adversarial Networks Combining Texture Loss) model which was also based on generative adversarial networks. We redefined the generator network and discriminator network. Firstly, on the network structure, residual dense blocks without excess batch normalization layers were used to form generator network. Visual Geometry Group (VGG)19 network was adopted as the basic framework of discriminator network. Secondly, in the loss function, the weighting of the four loss functions of texture loss, perceptual loss, adversarial loss and content loss was used as the objective function of generator. Texture loss was proposed to encourage local information matching. Perceptual loss was enhanced by employing the features before activation layer to calculate. Adversarial loss was optimized based on WGAN-GP (Wasserstein GAN with Gradient Penalty) theory. Content loss was used to ensure the accuracy of low-frequency information. During the optimization process, the target image information was reconstructed from different angles of high and low frequencies. Results: The experimental results showed that our method made the average Peak Signal to Noise Ratio of reconstructed images reach 27.99 dB and the average Structural Similarity Index reach 0.778 without losing too much speed, which was superior to other comparison algorithms in objective evaluation index. What is more, TSRGAN significantly improved subjective visual evaluations such as brightness information and texture details. We found that it could generate images with more realistic textures and more accurate brightness, which were more in line with human visual evaluation. Conclusions: Our improvements to the network structure could reduce the model’s calculation amount and stabilize the training direction. In addition, the loss function we present for generator could provide stronger supervision for restoring realistic textures and achieving brightness consistency. Experimental results prove the effectiveness and superiority of TSRGAN algorithm.

Sign in / Sign up

Export Citation Format

Share Document