Deep Generative Adversarial Networks Research Articles

A Deep Learning-Based Two-Branch Generative Adversarial Network for Image De-Raining

Raindrops can scatter and absorb light, causing images to become blurry or distorted. To improve image quality by reducing the impact of raindrops, this paper proposes a novel generative adversarial network for image de-raining. The network comprises two parts: a generative network and an adversarial network. The generative network performs image de-raining. The adversarial network determines whether the input image is rain-free or de-rained. The generative network comprises two branches: the A branch, which follows a traditional convolutional network structure, and the U branch, which utilizes a U-Net architecture. The A branch includes a multi-scale module for extracting information at different scales and a residual attention module to reduce redundant information interference. The U branch contains an encoder module designed to address the loss of details and local information caused by conventional down-sampling. To improve the performance of the generative network in image de-raining, this paper employs a relative discriminator incorporating a mean squared error loss. This discriminator better measures the differences between rainy and rain-free images while effectively preventing the occurrence of gradient vanishing. Finally, this study performs visual and quantitative comparisons of the proposed method and existing methods on three established rain image datasets. In the quantitative experiments, the proposed method outperforms existing methods regarding PSNR, SSIM, and VIF metrics. Specifically, our method achieves an average PSNR, SSIM, and VIF of approximately 5%, 3%, and 4% higher than the MFAA-GAN method, respectively. These results indicate that the de-rained images generated via the proposed method are closer to rain-free images.

Improving Non-Line-of-Sight Identification in Cellular Positioning Systems Using a Deep Autoencoding and Generative Adversarial Network Model

Positioning service is a critical technology that bridges the physical world with digital information, significantly enhancing efficiency and convenience in life and work. The evolution of 5G technology has proven that positioning services are integral components of current and future cellular networks. However, positioning accuracy is hindered by non-line-of-sight (NLoS) propagation, which severely affects the measurements of angles and delays. In this study, we introduced a deep autoencoding channel transform-generative adversarial network model that utilizes line-of-sight (LoS) samples as a singular category training set to fully extract the latent features of LoS, ultimately employing a discriminator as an NLoS identifier. We validated the proposed model in 5G indoor and indoor factory (dense clutter, low base station) scenarios by assessing its generalization capability across different scenarios. The results indicate that, compared to the state-of-the-art method, the proposed model markedly diminished the utilization of device resources and achieved a 2.15% higher area under the curve while reducing computing time by 12.6%. This approach holds promise for deployment in future positioning terminals to achieve superior localization precision, catering to commercial and industrial Internet of Things applications.

Electric load forecasting under false data injection attacks via denoising deep learning and generative adversarial networks

AbstractAccurate electric load forecasting at various time periods is considered a necessary challenge for electricity consumers and generators to maximize their economic efficiency in energy markets. Hence, the accuracy and effectiveness of existing electric load forecasting approaches depends on the data quality. Nowadays, with the implementation of modern power systems and Internet of Things technology, forecasting models are faced with a large volume of data, which puts the security and health of data at risk due to the use of numerous measuring devices and the threat of cyber‐attackers. In this study, a cyber‐resilient hybrid deep learning‐based model is developed that accurately forecasts electric load in short‐term and long‐term time horizons. The architecture of the proposed model systematically integrates stacked multilayer denoising autoencoder (SMDAE) and generative adversarial network (GAN) and is called SMDAE‐GAN. In the proposed framework, SMDAE layer is used to pre‐process and remove real fs and intentional anomalies in data, and GAN layer is also utilized to forecast electric load values. The effectiveness of the SMDAE‐GAN structure is studied using realistic electrical load data monitored in the distribution network of Tabriz, Iran, and meteorological data measured in weather station there. Compared with other conventional load forecasting approaches, the developed framework has the highest accuracy in both cases of using normal data with real‐world noise and damaged data under false data injection attacks.

An approach for two-dimensional convolutional neural networks for hourly passenger boarding demand prediction based on uneven smart-card data

An invaluable resource for understanding passenger boarding patterns and forecasting future travel demand is the tap-on smart-card data. Positive instances, on the other hand—boarding at a given bus stop at a certain time—are less common than negative instances when looking at the smart-card data (or instances) by boarding stops and by time of day. Machine learning algorithms that are used to estimate hourly boarding numbers at a certain location have been shown to be much less accurate when the data is imbalanced. Before using the smart-card data to forecast bus boarding demand, this research tackles the problem of data imbalance in the data. To create fake traveling instances to add into a synthetic training dataset containing more evenly distributed traveling and non-traveling examples, we suggest using deep generative adversarial networks (Deep-GAN). Next, a deep neural network, or DNN, is trained on the synthetic dataset to predict which instances from a given stop in a certain time frame will travel and which ones won't. According to the findings, resolving the data imbalance problem may greatly enhance the predictive model's functionality and make it more accurate in predicting ridership profiles. The suggested strategy may create a synthetic training set with a better similarity so diversity and, therefore, a stronger prediction capability, according to a comparison of the Deep-GAN's performance with other conventional resampling techniques. The study emphasizes the importance of the issue and offers helpful recommendations for enhancing the quality of the data and model performance for individual travel behavior analysis and travel behavior prediction.

Deep Generative Adversarial Networks For Noise Reduction in Medical Images: A Review

Deep Generative Adversarial Networks For Noise Reduction in Medical Images: A Review

Detection of real-time deep fakes and face forgery in video conferencing employing generative adversarial networks

As facial modification technology advances rapidly, it poses a challenge to methods used to detect fake faces. The advent of deep learning and AI-based technologies has led to the creation of counterfeit photographs that are more difficult to discern apart from real ones. Existing Deep fake detection systems excel at spotting fake content with low visual quality and are easily recognized by visual artifacts. The study employed a unique active forensic strategy Compact Ensemble-based discriminators architecture using Deep Conditional Generative Adversarial Networks (CED-DCGAN), for identifying real-time deep fakes in video conferencing. DCGAN focuses on video-deep fake detection on features since technologies for creating convincing fakes are improving rapidly. As a first step towards recognizing DCGAN-generated images, split real-time video images into frames containing essential elements and then use that bandwidth to train an ensemble-based discriminator as a classifier. Spectra anomalies are produced by up-sampling processes, standard procedures in GAN systems for making large amounts of fake data films. The Compact Ensemble discriminator (CED) concentrates on the most distinguishing feature between the natural and synthetic images, giving the generators a robust training signal. As empirical results on publicly available datasets show, the suggested algorithms outperform state-of-the-art methods and the proposed CED-DCGAN technique successfully detects high-fidelity deep fakes in video conferencing and generalizes well when comparing with other techniques. Python tool is used for implementing this proposed study and the accuracy obtained for proposed work is 98.23 %.

Building Face Ageing Model Using Face Synthesis

Advancements in face synthesis technology have enabled innovative methods for modeling facial aging. This research paper focuses primarily on creating a robust face aging model using deep learning and Generative Adversarial Networks (GANs), trained on a diverse dataset of facial images. The proposed approach captures both global features and local textures to produce realistic age-progressed images while preserving the subject's identity. This paper also examines face synthesis techniques, with specific emphasis for the various practical usage of GANs. The key objective of our project is to upgrade both the discriminator and the generator parts of GANs to generate more realistic, age- progressed face images. We evaluated the model using quantitative metrics and qualitative assessments, demonstrating its effectiveness. Additionally, we address ethical considerations, proposing guidelines for responsible use. Our study offers a novel framework with significant applications in security, forensics, and entertainment, and suggests future research directions to improve accuracy and ethical standards.

Data augmentation aided excavator activity recognition using deep convolutional conditional generative adversarial networks

Excavator activity recognition is a crucial task with significant practical applications such as assessing production efficiency, injury prevention, and enabling intelligent control. However, implementing deep learning-based activity recognition methods presents considerable challenges due to the requirement of large training datasets. And sensor-based activity recognition demands more convenient and reliable sensors. This paper proposes a generalized data augmentation framework for excavator activity recognition. Main pump pressure and flow signals are generated using Deep Convolutional Conditional Generative Adversarial Networks (DC-CGAN). The proposed framework integrates Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) architectures for activity recognition. Results indicate that the framework outperforms traditional data augmentation methods regarding the quality of generated data and shallow and single deep networks regarding model accuracy and generalization. Additionally, the framework was implemented on three sizes of excavators and a wheel loader, demonstrating excellent versatility.

SWIR based estimation of TIR emissivity of bare soil surfaces using deep conditional generative adversarial network in Landsat data

SWIR based estimation of TIR emissivity of bare soil surfaces using deep conditional generative adversarial network in Landsat data

Advancing ASD identification with neuroimaging: a novel GARL methodology integrating Deep Q-Learning and generative adversarial networks

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition that affects an individual's behavior, speech, and social interaction. Early and accurate diagnosis of ASD is pivotal for successful intervention. The limited availability of large datasets for neuroimaging investigations, however, poses a significant challenge to the timely and precise identification of ASD. To address this problem, we propose a breakthrough approach, GARL, for ASD diagnosis using neuroimaging data. GARL innovatively integrates the power of GANs and Deep Q-Learning to augment limited datasets and enhance diagnostic precision. We utilized the Autistic Brain Imaging Data Exchange (ABIDE) I and II datasets and employed a GAN to expand these datasets, creating a more robust and diversified dataset for analysis. This approach not only captures the underlying sample distribution within ABIDE I and II but also employs deep reinforcement learning for continuous self-improvement, significantly enhancing the capability of the model to generalize and adapt. Our experimental results confirmed that GAN-based data augmentation effectively improved the performance of all prediction models on both datasets, with the combination of InfoGAN and DQN's GARL yielding the most notable improvement.

New 2D to 3D Reconstruction of Heterogeneous Porous Media via Deep Generative Adversarial Networks (GANs)

AbstractAccurately characterizing rock microstructures in three dimensions (3D) is crucial for modeling various physical phenomena and estimating rock properties. Despite advancements in 3D imaging, limitations arise from the trade‐off between sample size and resolution, particularly in heterogeneous rocks with multi‐scale features where both high resolution and a large field of view are essential. These challenges have prompted interest in accurate 3D reconstructions from high‐resolution two‐dimensional (2D) images using advanced generative models like generative adversarial networks (GANs). In this study, using scanning electron microscopy and optical microscopy, we acquired 2D images from three orthogonal sections of a Berea sandstone sample. These images were employed to train a modified SliceGAN model, a variant of GANs, for 3D reconstruction. Unlike previous studies utilizing methods for 2D to 3D reconstructions that typically incorporated 3D images in their training, our approach relies exclusively on 2D images. We propose a systematic workflow which enables us to produce 3D reconstructions that closely mirror the original 2D inputs and a 3D X‐ray tomography in terms of structural and morphological characteristics. Additionally, we highlight the importance of input data size and training our model with representative images which enable us to generate diverse reconstructions with transport properties that align with previous studies on Berea sandstone. This underscores the potential of 2D to 3D reconstructions as an effective alternative to multiple X‐ray tomographies, integral for assessing variability in heterogeneous rocks.

Reinforced black widow algorithm with restoration technique based on optimized deep generative adversarial network

ABSTRACT Image restoration is used to develop the quality of image that is triggered by various noises and blurring. During this causes, certain areas in the images are vanished. The existing works does not provide sufficient restoration process with high accuracy. Therefore, a new image restoration system based on Optimized Deep Generative Adversarial Network (DGAN) with Reinforced Black Widow algorithm (BWOA) is proposed in this paper to increase the restoration accuracy and reducing the noises. At first, the input image is converted as gray scale image and the multi-scale edge information is removed as damaged area of an image by constructing a smooth function. Here, the extracted multi-scale edge information is given to the DGAN model. After that, the images are trained to create the best fake images through continuous play among generator and discriminator. Then, the detected images are restored in the original image with high accuracy. The hyper parameters of the DGAN are optimized by using the BWOA. The major objective of this paper is ‘to increase the restoration accuracy and the quality of the image by decreasing the noises occurred in the input image.’ The simulation process is performed on the MATLAB platform. The proposed DGAN-BWOA-IR attains higher restoration accuracy of 9.3%, higher PSNR value 74.589(db), SSIM 9.023% the proposed system is likened with the existing approaches, such as plug-and-play image restoration including deep denoiser prior (DCNN-IR), Learning enriched features for rapid image restoration with enhancement (LEF-IR), Exploiting deep generative prior for versatile image restoration with manipulation (GAN-IR), respectively.

Deep-KEDI: Deep learning-based zigzag generative adversarial network for encryption and decryption of medical images.

Medical imaging techniques have improved to the point where security has become a basic requirement for all applications to ensure data security and data transmission over the internet. However, clinical images hold personal and sensitive data related to the patients and their disclosure has a negative impact on their right to privacy as well as legal ramifications for hospitals. In this research, a novel deep learning-based key generation network (Deep-KEDI) is designed to produce the secure key used for decrypting and encrypting medical images. Initially, medical images are pre-processed by adding the speckle noise using discrete ripplet transform before encryption and are removed after decryption for more security. In the Deep-KEDI model, the zigzag generative adversarial network (ZZ-GAN) is used as the learning network to generate the secret key. The proposed ZZ-GAN is used for secure encryption by generating three different zigzag patterns (vertical, horizontal, diagonal) of encrypted images with its key. The zigzag cipher uses an XOR operation in both encryption and decryption using the proposed ZZ-GAN. Encrypting the original image requires a secret key generated during encryption. After identification, the encrypted image is decrypted using the generated key to reverse the encryption process. Finally, speckle noise is removed from the encrypted image in order to reconstruct the original image. According to the experiments, the Deep-KEDI model generates secret keys with an information entropy of 7.45 that is particularly suitable for securing medical images.

Image Processing and Optimization Using Deep Learning-Based Generative Adversarial Networks (GANs)

Image Processing and Optimization Using Deep Learning-Based Generative Adversarial Networks (GANs)

CNNFDNF: CONVOLUTIONAL NEURAL NETWORK AND CASCADE NEURO FUZZY NETWORK FOR SEVERITY LEVEL OF COVID-19 DETECTION USING CHEST X-RAY IMAGES

Chest X-ray (CXR) images are the most important imaging modality to monitor and detect several lung diseases and Coronavirus Disease 2019 (COVID-19) disease. This is the first imaging method in the detection of COVID-19. Since the limited availability of interpreted medical images, and the accuracy are the biggest challenges in the conventional methods. Hence, a novel framework is designed for the severity level of COVID-19 detection using CXR images named Convolutional Neural Network and Cascade Neuro-Fuzzy Network (CNNFDNF). Initially, the input CXR image is allowed to the preprocessing phase that is done utilizing the Kalman filter and Region of Interest (ROI) extraction. Then, the lung lobe segmentation is executed by Psi-Net. After that, the extraction of Convolutional Neural Network (CNN) features is conducted from the interested lobe region. Moreover, the extracted image features are subjected to COVID-19 detection, where it is achieved by Deep Generative Adversarial Network (Deep GAN) trained by Henry Gas Water Wave Optimization (HGWWO). Here, HGWWO is blended by Water Wave Optimization (WWO) and Henry Gas Solubility Optimization (HGSO). The detection phase is categorized into COVID and non-COVID. If the detected image is COVID, then the CXR images are given to the second level classification. Finally, it is accomplished by employing the proposed CNNFDNF to classify them into low, moderate, or high COVID-19 infection affected at lungs. The proposed CNNFDNF is the newly devised scheme by modifying the layers of CNN and cascaded neuro-fuzzy network (DNF) structure. The measures employed for this scheme namely, accuracy, sensitivity and specificity acquired are 93.0%, 94.1%, and 93.8%. The proposed method is essential for keeping the situation under control. Because, in the severe cases may have severe pneumonia, other organ failure, and possible death.

A review of deep learning and Generative Adversarial Networks applications in medical image analysis

Nowadays, computer-aided decision support systems (CADs) for the analysis of images have been a perennial technique in the medical imaging field. In CADs, deep learning algorithms are widely used to perform tasks like classification, identification of patterns, detection, etc. Deep learning models learn feature representations from images rather than handcrafted features. Hence, deep learning models are quickly becoming the state-of-the-art method to achieve good performances in different computer-aided decision-support systems in medical applications. Similarly, deep learning-based generative models called Generative Adversarial Networks (GANs) have recently been developed as a novel method to produce realistic-looking synthetic data. GANs are used in different domains, including medical imaging generation. The common problems, like class imbalance and a small dataset, in healthcare are well addressed by GANs, and it is a leading area of research. Segmentation, reconstruction, detection, denoising, registration, etc. are the important applications of GANs. So in this work, the successes of deep learning methods in segmentation, classification, cell structure and fracture detection, computer-aided identification, and GANs in synthetic medical image generation, segmentation, reconstruction, detection, denoising, and registration in recent times are reviewed. Lately, the review article concludes by raising research directions for DL models and GANs in medical applications.

Distance Regression Enhanced with Temporal Information Fusion and Adversarial Training for Robot-Assisted Endomicroscopy.

Probe-based confocal laser endomicroscopy (pCLE) has a role in characterising tissue intraoperatively to guide tumour resection during surgery. To capture good quality pCLE data which is important for diagnosis, the probe-tissue contact needs to be maintained within a working range of micrometre scale. This can be achieved through micro-surgical robotic manipulation which requires the automatic estimation of the probe-tissue distance. In this paper, we propose a novel deep regression framework composed of the Deep Regression Generative Adversarial Network (DR-GAN) and a Sequence Attention (SA) module. The aim of DR-GAN is to train the network using an enhanced image-based supervision approach. It extents the standard generator by using a well-defined function for image generation, instead of a learnable decoder. Also, DR-GAN uses a novel learnable neural perceptual loss which combines for the first time spatial and frequency domain features. This effectively suppresses the adverse effects of noise in the pCLE data. To incorporate temporal information, we've designed the SA module which is a cross-attention module, enhanced with Radial Basis Function based encoding (SA-RBF). Furthermore, to train the regression framework, we designed a multi-step training mechanism. During inference, the trained network is used to generate data representations which are fused along time in the SA-RBF module to boost the regression stability. Our proposed network advances SOTA networks by addressing the challenge of excessive noise in the pCLE data and enhancing regression stability. It outperforms SOTA networks applied on the pCLE Regression dataset (PRD) in terms of accuracy, data quality and stability.

A deep learning approach for prediction of air quality index in smart city

Industrial developments and consumption of massive amount of fossil fuels, vehicle pollution, and other calamities upsurges the AQI (Air Quality Index) of major cities in a drastic manner. Owing to these factors, it is important to take proactive measures for reducing the air pollution in order to avoid life- threatening consequence. Therefore, prediction of air quality is significant for improving the health of living beings as highly polluted regions have a higher concentration of pollutants mixed in the air, affecting the respiratory system and reducing the lifetime. To control pollution, AQI is used as a measure for estimating the pollutant content in the air. Even though many existing techniques have predicted AQI, enhancement is required in prediction algorithms with minimized loss. To address the challenges in traditional algorithms, the proposed smart cities-based AQI prediction intends to utilize the proposed regression algorithm in the dataset, namely Air- Quality-Data, which collected harmful pollutants on an hourly and daily basis from multiple cities in India between 2015 to 2020. To achieve prediction efficiency with reduced loss, pre-processing of input data is being performed using Deep GAN (Generative Adversarial Network). It performs the imputation of data in place of missing values to improve accurate prediction. Additionally, feature scaling normalizes independent real-data features to a fixed scale. With the processed data, regression is done through modified Stacked Attention GRU with KL divergence, which predicts Ernakulam, Chennai and Ahmedabad cities with higher, medium, and low levels of AQI in India. The performance of the proposed regression algorithm is measured using metrics such as MAE (Mean Absolute Error), MSE (Mean Square Error), R2 (Coefficient of determination), MAPE (Mean Absolute Percentage Error), and RMSE (Root Mean Square Error) and better MAE, MSE, R2, MAPE and RMSE obtained by the model is 0.1013, 0.0134, 0.9479, 0.1152 and 0.1156. Internal assessment and comparative analysis performed with existing regression algorithms exhibit lower loss values obtained from the present research, which determines the efficacy of the proposed model.

Compressed Video Sensing Based on Deep Generative Adversarial Network

Compressed Video Sensing Based on Deep Generative Adversarial Network

Autism Spectrum Disorder Detection Using Parallel Deep Convolution Neural Network and Generative Adversarial Networks

Autism Spectrum Disorder Detection Using Parallel Deep Convolution Neural Network and Generative Adversarial Networks