Enhancement Generative Adversarial Network Research Articles

Semi-Supervised DEGAN for Optical High-Resolution Remote Sensing Image Scene Classification

Semi-supervised methods have made remarkable achievements via utilizing unlabeled samples for optical high-resolution remote sensing scene classification. However, the labeled data cannot be effectively combined with unlabeled data in the existing semi-supervised methods during model training. To address this issue, we present a semi-supervised optical high-resolution remote sensing scene classification method based on Diversity Enhanced Generative Adversarial Network (DEGAN), in which the supervised and unsupervised stages are deeply combined in the DEGAN training. Based on the unsupervised characteristic of the Generative Adversarial Network (GAN), a large number of unlabeled and labeled images are jointly employed to guide the generator to obtain a complete and accurate probability density space of fake images. The Diversity Enhanced Network (DEN) is designed to increase the diversity of generated images based on massive unlabeled data. Therefore, the discriminator is promoted to provide discriminative features by enhancing the generator given the game relationship between two models in DEGAN. Moreover, the conditional entropy is adopted to make full use of the information of unlabeled data during the discriminator training. Finally, the features extracted from the discriminator and VGGNet-16 are employed for scene classification. Experimental results on three large datasets demonstrate that the proposed scene classification method yields a superior classification performance compared with other semi-supervised methods.

Integrated Framework for Intrusion Detection Through Adversarial Sampling and Enhanced Deep Correlated Hierarchical Network

Intrusion Detection Systems (IDSs) play a critical role in detecting malicious assaults and threats in the network system. This research work proposed a network intrusion detection technique, which combines an Adversarial Sampling and Enhanced Deep Correlated Hierarchical Network for IDS. Initially, the proposed Enhanced Generative Adversarial Networks (EGAN) method is used to raise the minority sample. A balanced dataset can be created in this way, allowing the model to completely learn the properties of minority samples while also drastically minimizing the model training time. Then, create an Enhanced Deep Correlated Hierarchical Network model by using a Bi-Directional Long Short-Term Memory (BiLSTM) to collect temporal characteristics and Cross-correlated Convolution Neural Network (CCNN) to retrieve spatial characteristics. The softmax classifier at the end of BiLSTM is used to classify intrusion data. The traditional NSL-KDD dataset is utilized for the experimentation of the proposed model.

Single Image Super-Resolution Method Based on an Improved Adversarial Generation Network

Super-Resolution (SR) techniques for image restoration have recently been gaining attention due to their excellent performance. For powerful learning abilities, Generative Adversarial Networks (GANs) have been proven to have achieved great success. In this paper, we propose an Enhanced Generative Adversarial Network (EGAN) for improving its effects for a real-time Super-Resolution task. The main content of this paper are as follows: (1) We adopted the Laplacian pyramid framework as a pre-trained module, which is beneficial for providing multiscale features for our input. (2) At each feature block, a convolutional skip-connections network, which may contain some latent information, was significant for the generative model to reconstruct a plausible-looking image. (3) Considering that the edge details usually play an important role in image generation, a perceptual loss function was defined to train and seek the optimal parameters. Quantitative and qualitative evaluations were demonstrated so that our algorithm not only took full advantage of the Convolutional Neural Networks (CNNs) to improve the image quality, but also performed better than other algorithms in speed and performance for real-time Super-Resolution tasks.

Rolling Bearing Fault Diagnosis in Limited Data Scenarios Using Feature Enhanced Generative Adversarial Networks

Most current rolling bearing fault diagnostic approaches presume that sufficient labeled data are available for training. However, labeled fault data collection in realistic engineering is time-consuming and expensive, whereas the majority of data available are unlabeled, restricting diagnostic performance. To alleviate the dependence of feature self-extraction model on rolling bearing fault data and improve the feature capture ability, a bearing fault diagnosis method based on feature-enhanced generative adversarial networks with an auxiliary classifier (AC-FEGAN) is proposed in this paper. Firstly, the non-saturating loss with gradient penalty term is introduced to circumvent the gradient vanishing and improve learning stability. Then, employing identity mapping of residual networks introduced into AC-FEGAN, representative faulty features can be automatically extracted, overcoming the restrictions of shallow convolutional neural networks in feature extraction. Simultaneously, to improve feature learning ability, the self-attention module embedded in the residual network accomplishes refined feature extraction by incorporating appropriate weight matrices into the faulty feature maps. Finally, the auxiliary classifier is employed as a pre-training model for the fault diagnosis model, with generated and unlabeled samples employed to fine-tune the auxiliary classifier to identify faulty bearings. The experimental results show that the proposed method can significantly improve the quality of generated samples while eliminating the reliance of existing diagnosis methods on data volume to achieve more accurate and effective fault diagnosis.

FSPBO-DQN: SeGAN based segmentation and Fractional Student Psychology Optimization enabled Deep Q Network for skin cancer detection in IoT applications

Skin cancer is one of the dangerous types of cancer and the rate of death is increasing due to the lack of knowledge in prevention and the symptoms. It is a common cancer type around the world and it occurs when the skin cells are damaged. Hence, the detection of skin cancer near the beginning is important to prevent the spread of cancer and to increase the survival rate. Recently, image processing and machine learning techniques gained more interest in medical applications. However, early analysis of skin cancer images is very challenging due to factors, like variations in the color illumination, light reflections from the skin surface, and different sizes and shapes of lesions. To detect skin cancer at an early stage and to increase the survival rate, an effective skin cancer detection method is introduced in this study using the proposed Fractional Student Psychology Based Optimization-based Deep Q Network (FSPBO-based DQN) in the wireless network scenario. At first, the nodes simulated in the network area are allowed to capture the healthcare information to make the detection strategy using the proposed method. Then, the routing is performed by the proposed Fractional Student Psychology Based Optimization (FSPBO) algorithm by considering the fitness parameters, like distance, energy, trust, and delay. After the images (healthcare information) are reached the Base Station (BS), the pre-processing, segmentation, and cancer detection processes are carried out to detect the skin lesions. Initially, the image is fed to pre-processing phase, where a Type II Fuzzy System and cuckoo search optimization algorithm (T2FCS) filter is employed to remove the noise of images. Then, the pre-processed images are fed to the segmentation phase, where speech enhancement Generative Adversarial Network (SeGAN) is used to generate the segmented results. Afterward, the Deep Q Network (DQN) detects the skin cancer based on the segmented results, and the training of DQN is made using the proposed FSPBO algorithm, which is designed by integrating the Student Psychology Based Optimization (SPBO) and Fractional Calculus (FC). The proposed method is more robust and reduces computation time and complexity. Moreover, the proposed method achieved higher performance by considering the measures, namely accuracy, sensitivity, and specificity with the values of 92.364%, 93.20%, and 92.63%.

Enhanced Generative Adversarial Networks with Restart Learning Rate in Discriminator

A series of Generative Adversarial Networks (GANs) could effectively capture the salient features in the dataset in an adversarial way, thereby generating target data. The discriminator of GANs provides significant information to update parameters in the generator and itself. However, the discriminator usually becomes converged before the generator has been well trained. Due to this problem, GANs frequently fail to converge and are led to mode collapse. This situation can cause inadequate learning. In this paper, we apply restart learning in the discriminator of the GAN model, which could bring more meaningful updates for the training process. Based on CIFAR-10 and Align Celeba, the experiment results show that the proposed method could improve the performance of a DCGAN with a low FID score over a stable learning rate scheme. Compared with two other stable GANs—SNGAN and WGAN-GP—the DCGAN with a restart schedule had a satisfying performance. Compared with the Two Time-Scale Update Rule, the restart learning rate is more conducive to the training of DCGAN. The empirical analysis indicates four main parameters have varying degrees of influence on the proposed method and present an appropriate parameter setting.

Lightweight End-to-End Speech Enhancement Generative Adversarial Network Using Sinc Convolutions

Generative adversarial networks (GANs) have recently garnered significant attention for their use in speech enhancement tasks, in which they generally process and reconstruct speech waveforms directly. Existing GANs for speech enhancement rely solely on the convolution operation, which may not accurately characterize the local information of speech signals—particularly high-frequency components. Sinc convolution has been proposed in order to allow the GAN to learn more meaningful filters in the input layer, and has achieved remarkable success in several speech signal processing tasks. Nevertheless, Sinc convolution for speech enhancement is still an under-explored research direction. This paper proposes Sinc–SEGAN, a novel generative adversarial architecture for speech enhancement, which usefully merges two powerful paradigms: Sinc convolution and the speech enhancement GAN (SEGAN). There are two highlights of the proposed system. First, it works in an end-to-end manner, overcoming the distortion caused by imperfect phase estimation. Second, the system derives a customized filter bank, tuned for the desired application compactly and efficiently. We empirically study the influence of different configurations of Sinc convolution, including the placement of the Sinc convolution layer, length of input signals, number of Sinc filters, and kernel size of Sinc convolution. Moreover, we employ a set of data augmentation techniques in the time domain, which further improve the system performance and its generalization abilities. Compared to competitive baseline systems, Sinc–SEGAN overtakes all of them with drastically reduced system parameters, demonstrating its effectiveness for practical usage, e.g., hearing aid design and cochlear implants. Additionally, data augmentation methods further boost Sinc–SEGAN performance across classic objective evaluation criteria for speech enhancement.

Enhanced generative adversarial network for extremely imbalanced fault diagnosis of rotating machine

Fault diagnosis is the key procedure to ensure the stability and reliability of mechanical equipment operation. Recent works show that deep learning-based methods outperform most of traditional fault diagnosis techniques. However, a practical problem comes up in these studies, where deep learning models cannot be well trained and the classification accuracy is greatly affected because of the sample-imbalance problem, which means that there are a large amount of normal data but few fault samples. To solve the problem, an enhanced generative adversarial network (E-GAN) is proposed. Firstly, the deep convolutional generative adversarial network (DCGAN) is utilized to generate more samples to balance the training set. Then, by integrating K-means clustering algorithm, we developed a modified CNN diagnosis model for fault classification. The experiment results demonstrate that the proposed E-GAN can greatly improve the classification accuracy and is superior to the compared methods.

TAU-Net: Temporal Activation U-Net Shared With Nonnegative Matrix Factorization for Speech Enhancement in Unseen Noise Environments

In this paper, a novel speech enhancement method based on a hybrid machine-learning architecture consisting of U-Net and nonnegative matrix factorization (NMF) is proposed. The proposed method attempts to take advantage of both the accurate separation for known noise environments by U-Net and the adaptation to unseen noises by an NMF with an online dictionary learning technique. To merge the two different architectures, a modified U-Net with a temporal activation layer (TAU-Net) is jointly optimized with NMF models that represent universal speech and noise. The proposed method first estimates the temporal activations from the encoder of the proposed TAU-Net. Then, an NMF with online dictionary learning adjusts the initially given temporal activations to suppress their cross-activations due to unseen noises that are unknown in the training phase of TAU-Net. Finally, clean speech is obtained by adjusting temporal activations to the TAU-Net decoder. The effectiveness of the proposed TAU-Net-based speech enhancement method is evaluated in various unseen noise environments. Consequently, the proposed method achieves a substantial improvement with average signal-to-distortion ratios of 2.32 dB and 5.68 dB, which are higher than those of the baseline methods such asspeech enhancement generative adversarial network (SEGAN) and U-Net, respectively.

Joint image-to-image translation with denoising using enhanced generative adversarial networks

Impressive progress has been made recently in image-to-image translation using generative adversarial networks (GANs). However, existing methods often fail in translating source images with noise to target domain. To address this problem, we joint image-to-image translation with image denoising and propose an enhanced generative adversarial network (EGAN). In particular, built upon pix2pix, we introduce residual blocks in the generator network to capture deeper multi-level information between source and target image distribution. Moreover, a perceptual loss is proposed to enhance the performance of image-to-image translation. As demonstrated through extensive experiments, our proposed EGAN can alleviate effects of noise in source images, and outperform other state-of-the-art methods significantly. Furthermore, we experimentally indicate that the proposed EGAN is also effective when applied to image denoising.

CEGAN: Classification Enhancement Generative Adversarial Networks for unraveling data imbalance problems

The data imbalance problem in classification is a frequent but challenging task. In real-world datasets, numerous class distributions are imbalanced and the classification result under such condition reveals extreme bias in the majority data class. Recently, the potential of GAN as a data augmentation method on minority data has been studied. In this paper, we propose a classification enhancement generative adversarial networks (CEGAN) to enhance the quality of generated synthetic minority data and more importantly, to improve the prediction accuracy in data imbalanced condition. In addition, we propose an ambiguity reduction method using the generated synthetic minority data for the case of multiple similar classes that are degenerating the classification accuracy. The proposed method is demonstrated with five benchmark datasets. The results indicate that approximating the real data distribution using CEGAN improves the classification performance significantly in data imbalanced conditions compared with various standard data augmentation methods.

DE-GAN: A Conditional Generative Adversarial Network for Document Enhancement.

Documents often exhibit various forms of degradation, which make it hard to be read and substantially deteriorate the performance of an OCR system. In this paper, we propose an effective end-to-end framework named document enhancement generative adversarial networks (DE-GAN) that uses the conditional GANs (cGANs) to restore severely degraded document images. To the best of our knowledge, this practice has not been studied within the context of generative adversarial deep networks. We demonstrate that, in different tasks (document clean up, binarization, deblurring and watermark removal), DE-GAN can produce an enhanced version of the degraded document with a high quality. In addition, our approach provides consistent improvements compared to state-of-the-art methods over the widely used DIBCO 2013, DIBCO 2017, and H-DIBCO 2018 datasets, proving its ability to restore a degraded document image to its ideal condition. The obtained results on a wide variety of degradation reveal the flexibility of the proposed model to be exploited in other document enhancement problems.

Improving generative adversarial networks for speech enhancement through regularization of latent representations

Speech enhancement aims to improve the quality and intelligibility of speech signals, which is a challenging task in adverse environments. Speech enhancement generative adversarial network (SEGAN) that adopted a generative adversarial network (GAN) for speech enhancement achieved promising results. In this paper, a new network architecture and loss function based on SEGAN are proposed for speech enhancement. Different from most network structures applied in this field, the new network, called high-level GAN (HLGAN), uses parallel noisy and clean speech signals as input in the training phase instead of only noisy speech signals, which enables us to make full use of the information carried by the clean speech signals. Additionally, we introduce a new supervised speech representation loss, also known as high-level loss, in the middle hidden layer of the generative network. The high-level loss function is advantageous to HLGAN in speech enhancement under low signal-to-noise (SNR) environments and low-resource environments. We evaluate the performance of HLGAN over a wide range of experiments, in which our model produces significant improvements. Extensive experiments further demonstrate the generality of our model in a variety of speech enhancement cases. The issue of SEGAN losing speech components while removing noise in low SNR environments is improved. In addition, HLGAN can effectively enhance the speech signals of two low-resource languages simultaneously. The reasons for the superior performance of HLGAN are discussed.

Imbalanced Fault Diagnosis of Rolling Bearing Using Enhanced Generative Adversarial Networks

Machinery fault diagnosis tasks have been well addressed when sufficient and abundant data are available. However, the data imbalance problem widely exists in real-world scenarios, which leads to the performance deterioration of fault diagnosis markedly. To solve this problem, we present a novel imbalanced fault diagnosis method based on the enhanced generative adversarial networks (GAN). By artificially generating fake samples, the proposed method can mitigate the loss caused by the lack of real fault data. Specifically, in order to improve the quality of generated samples, a new discriminator is designed using spectrum normalization (SN) strategy and a two time-scale update rule (TTUR) method is used to stabilize the training process of GAN. Then, an enhanced Wasserstein GAN with gradient penalty is developed to generate high-quality synthetic samples for the fault samples set. Finally, a deep convolutional classifier is constructed to carry out fault classification. The performance and effectiveness of the proposed method are validated on the Case Western Reserve University bearing dataset and rolling bearing dataset acquired from our laboratory. The simulation results show that the proposed method has a superior performance than other methods for imbalanced fault diagnosis tasks.

Towards Unsupervised Deep Image Enhancement with Generative Adversarial Network.

Improving the aesthetic quality of images is challenging and eager for the public. To address this problem, most existing algorithms are based on supervised learning methods to learn an automatic photo enhancer for paired data, which consists of low-quality photos and corresponding expert-retouched versions. However, the style and characteristics of photos retouched by experts may not meet the needs or preferences of general users. In this paper, we present an unsupervised image enhancement generative adversarial network (UEGAN), which learns the corresponding image-to-image mapping from a set of images with desired characteristics in an unsupervised manner, rather than learning on a large number of paired images. The proposed model is based on single deep GAN which embeds the modulation and attention mechanisms to capture richer global and local features. Based on the proposed model, we introduce two losses to deal with the unsupervised image enhancement: (1) fidelity loss, which is defined as a l2 regularization in the feature domain of a pre-trained VGG network to ensure the content between the enhanced image and the input image is the same, and (2) quality loss that is formulated as a relativistic hinge adversarial loss to endow the input image the desired characteristics. Both quantitative and qualitative results show that the proposed model effectively improves the aesthetic quality of images.

Enhanced generative adversarial networks for fault diagnosis of rotating machinery with imbalanced data

Deep learning-based methods have attracted the attention of researchers due to their outstanding performance in automatic feature learning, a crucial step for satisfactory fault diagnosis. However, faults in rotating machinery may occur occasionally, and fault-related signals are difficult to collect, resulting in imbalanced data. This problem is a major concern in fault diagnosis research. In this study, an enhanced generative adversarial network (EGAN) is proposed to establish a fault diagnosis model for rotating machinery. The model, based on a 2D convolutional network, consists of a generator and a discriminator. The generator produces specified samples to automatically enrich small samples for balancing datasets. The discriminator validates the distribution similarity between the generated and original samples. Fault types are recognized by a classifier using the generated and original samples. An adaptive training ratio strategy is also proposed to improve the convergence rapidity and stability of the EGAN training process. In case studies, two datasets are applied to verify the generalization performance of the EGAN. Results confirm the superior performance of the EGAN in fault diagnosis, particularly in the analyses of datasets with imbalanced data. Moreover, visualization results demonstrate that the proposed EGAN exhibits satisfactory ability and great potential for fault diagnosis applications of rotating machinery.

Interference cancellation algorithm based on improved generative model combined with improved generalised sidelobe cancellation in ELF communication

To improve the communication quality of extremely-low-frequency (ELF) communication effectively, an interference cancellation algorithm based on the generative model widely used in the artificial intelligence field is proposed. Some magnetic antennas with higher sensitivity and analogue circuits with lower noise floor are designed to suppress various out-of-band interferences. For the first time, a generative model is introduced into the interference cancellation field in ELF communication. Based on the speech enhancement generative adversarial network widely used in speech signal enhancement, an improved generative model that can be applied in the interference cancellation field is proposed to provide some more revelant reference information. By the combination of the improved generative model and the improved generalised sidelobe cancellation (GSC) algorithm, the estimated accuracy of noise and interference is improved effectively. To verify the effectiveness of the proposed algorithm, an experimental platform is built in a laboratory environment, and multiple sets of controlled experiments are performed. Also, the experimental results show that the improved generative model has better performance, better robustness, and relatively lower computational complexity. In addition, compared with the original GSC algorithm and its traditional improved algorithm, the proposed algorithm further promoted the signal-to-interference-plus-noise ratio gain within the range of signal bandwidth.

Defect Enhancement Generative Adversarial Network for Enlarging Data Set of Microcrack Defect

This paper presents a micro defect data set expansion method focuses on the microcrack defect of magnetic ring. Deep neural networks require a mass of training samples to be fully optimized. However, it is difficult to obtain a mass of defective samples in industrial field. In the case of insufficient samples, using GANs (Generative Adversarial Networks) for data expansion can effectively solve the problems of model over-fitting and low detection accuracy caused by insufficient training samples. However, it is difficult for conventional GANs to generate microcrack defective samples of high quality. This paper presents Defect Enhancement Generative Adversarial Network (DEGAN). This model can generate microcrack defects with obvious defect characteristics and high diversity. The experimental results show that the defective samples generated by DEGAN are very close to the real ones. The data set amplified by this model can significantly optimize deep neural network and achieve higher defect detection accuracy.