Deep Generative Adversarial Networks Research Articles

Circularly Polarized Fabry–Pérot Cavity Sensing Antenna Design Using Generative Model

We consider the problem of designing a circularly polarized Fabry-P'erot cavity (FPC) antenna for S-band sensing applications such as satellite navigation and communication. The spatial distribution of peripheral roughness of the unit cell of FPC's partially reflecting surface (PRS) serves as an important design optimization criterion. However, the evaluation of each candidate design using a full-wave solver is computationally expensive. To this end, we propose a deep generative adversarial network (GAN) for realizing a surrogate model that is trained with input-output pairs of antenna designs and their corresponding patterns. Using the GAN framework, we quickly evaluate the characteristics of a large volume of candidate designs and choose the antenna design with an axial ratio of 0.4 dB, a gain of 7.5 dB, and a bandwidth of 269 MHz.

Quantifying microstructures of earth materials using higher-order spatial correlations and deep generative adversarial networks

The key to most subsurface processes is to determine how structural and topological features at small length scales, i.e., the microstructure, control the effective and macroscopic properties of earth materials. Recent progress in imaging technology has enabled us to visualise and characterise microstructures at different length scales and dimensions. However, one limitation of these technologies is the trade-off between resolution and sample size (or representativeness). A promising approach to this problem is image reconstruction which aims to generate statistically equivalent microstructures but at a larger scale and/or additional dimension. In this work, a stochastic method and three generative adversarial networks (GANs), namely deep convolutional GAN (DCGAN), Wasserstein GAN with gradient penalty (WGAN-GP), and StyleGAN2 with adaptive discriminator augmentation (ADA), are used to reconstruct two-dimensional images of two hydrothermally rocks with varying degrees of complexity. For the first time, we evaluate and compare the performance of these methods using multi-point spatial correlation functions—known as statistical microstructural descriptors (SMDs)—ultimately used as external tools to the loss functions. Our findings suggest that a well-trained GAN can reconstruct higher-order, spatially-correlated patterns of complex earth materials, capturing underlying structural and morphological properties. Comparing our results with a stochastic reconstruction method based on a two-point correlation function, we show the importance of coupling training/assessment of GANs with higher-order SMDs, especially in the case of complex microstructures. More importantly, by quantifying original and reconstructed microstructures via different GANs, we highlight the interpretability of these SMDs and show how they can provide valuable insights into the spatial patterns in the synthetic images, allowing us to detect common artefacts and failure cases in training GANs.

Using RF Signals to Generate Indoor Maps

Generating maps of indoor environments beyond the line-of-sight finds applications in several areas such as planning, navigation, and security. While researchers have previously explored the use of RF signals to generate maps, prior work has two important limitations: (i) it requires moving the mapping setup along the entire lengths of the sides of the building, and (ii) it generates maps that are not fully connected, rather are scatter plots of locations from where some obstacles reflected the signals. Thus, prior approaches require human interpretation to locate the walls and determine how they merge. In this article, we address these limitations and propose RFMap, which generates fully connected maps, and does not require the measurement setup to be moved along the sides of the buildings. To generate the map, RFMap first transmits RF signals in many different directions and then measures the distances of different reflectors inside the building. Next, it identifies these reflectors and classifies them into various types based on the properties of the reflections. A key challenge is that RFMap does not receive reflections from all the directions due to the specular nature of the reflectors. Due to this, it only gets sparse data about the objects in the environment. To address this challenge, RFMap trains a deep generative adversarial network (GAN) to intelligently predict the missing information. At runtime, it feeds the locations and types of the detected reflectors to the trained GAN and generates complete and accurate map. We implemented RFMap using software defined radios and extensively evaluated it in several real-world environments. Our results show that RFMap generated the maps of all the buildings that we tested it on with high accuracy.

Wide & deep generative adversarial networks for recommendation system

Generative Adversarial Networks (GANs) has achieved great success in computer vision like Image Inpainting, Image Super-Resolution. Many researchers apply it to improve the effectiveness of recommendation system. However, GANs-based methods obtain users’ preferences using a single Neural Network framework in generative model, which may not be fully mined. Furthermore, most GANs-based algorithms adopt cross-entropy loss to get pair-wise bias, but these methods don’t reveal global data distribution loss when data are sparse. Those problems will influence the performance of the algorithm and result in poor accuracy. To address these problems, we introduce Wide & Deep Generative Adversarial Networks for Recommendation System (a.k.a W & DGAN) in this paper. On the one hand, we employ Wide & Deep Learning as a generative model capable of extracting both explicit and implicit information of user preferences. Furthermore, we combine Cross-Entropy loss in G with Wasserstein loss in D to get data distribution, then, the joint loss will be to receive the training information feedback from data distribution. Empirical results on three public benchmarks show that W&DGAN significantly outperforms state-of-the-art methods.

Landslide identification using water cycle particle swarm optimization‐based deep generative adversarial network

SummaryA natural vulnerability called a landslide threatens people, property, and infrastructure in many different places all over the world. Various landslide identification techniques are developed to assess the landslide hazard, but accurate identification of landslide occurrence particularly in certain regions remains a challenging issue, due to the complex scale‐dependent processes among geomorphometric features and landslides. To solve the shortcomings of landslide identification, the water cycle particle swarm optimization‐based deep generative adversarial network (WCPSO‐based Deep GAN) hybridization technique is presented. To develop the developed WCPSO, the water cycle algorithm (WCA) and particle swarm optimization (PSO) were hybridized. The landslide rainfall dataset is used as the source of the input data in this instance, and the Manhattan distance approach is then used to do feature selection. After performing data augmentation based on the results of the feature selection, the proposed WCPSO‐based deep GAN approach accurately detects the landslides. The developed WCPSO‐driven deep GAN's performance is also measured using the three metrics accuracy, specificity, and sensitivity, with a maximum accuracy of 0.937, higher specificity of 0.919, and higher sensitivity of 0.952.

Imbalanced Sample Fault Diagnosis of Rolling Bearing Using Deep Condition Multidomain Generative Adversarial Network

Rolling bearing fault diagnosis plays a crucial role in ensuring the safety and reliability of rotary machine, and some methods that employ deep-learning technology are grounded on the assumption that there are large-scale failure data. In reality, it is difficult to acquire more failure rolling bearing data under the different failure rates. In this article, a novel condition multidomain generative adversarial network (CMDGAN) is introduced for imbalanced sample fault diagnosis. This framework effectively captures the sample distribution information by a fusion of two-domain information when there are limited raw samples. Also, the introduced self-adaptive sample condition is no prior knowledge needed and contributes to the different state of synthetic data. Finally, an improved fault diagnostic model with the self-attention module implements the fusion of local fault feature and global periodic fault feature. Multiple sets of experiments on the Case Western Reserve University (CWRU) and Dalian University of Technology Vibration Lab (DUT-VL) datasets reveal the efficiency of our proposed approach, and the generating samples improve the fault diagnosis performance, which outperforms the contrasting methods.

Adaptive Coarse-to-Fine Single Pixel Imaging With Generative Adversarial Network Based Reconstruction

Single Pixel Imaging (SPI) that only uses one light intensity sensor has been researched extensively as an alternative imaging method. It is proven to work at low light conditions and unconventional wavelength bands where the classical pixel array sensors are limited. However, the major issues of SPI remain as the image quality and computational efficiency. Thus, this paper proposes an adaptive coarse-to-fine (C2F) sampling method to replace the typical uniform sampling method to achieve image reconstruction with better quality. This scalable sampling mechanism is adaptive to the target scene as it will progressively sample according to the image complexity and quality indicator. Subsequently, a deep Generative Adversarial Network (GAN) model is also proposed to improve the time efficiency of the multi-scale image reconstruction. The results show that C2F sampling consistently outperforms uniform sampling in terms of image quality (21% in SSIM, 8% in PSNR and 17% in RMSE). Besides, improvement in the efficiency is also achieved by the proposed GAN reconstruction, whereby the total time taken is only 0.025% of the time taken for the conventional L1 reconstruction method (≈ 4000 times faster). In conclusion, the proposed adaptive C2F SPI using GAN reconstruction method can serve as an optimised solution to improve both the image quality and computational efficiency in SPI.

Image Matting Algorithm Using Deep Generative Adversarial Networks

Image Matting Algorithm Using Deep Generative Adversarial Networks

Multi-Level Deep Generative Adversarial Networks for Brain Tumor Classification on Magnetic Resonance Images

The brain tumor is an abnormal and hysterical growth of brain tissues, and the leading cause of death affected patients worldwide. Even in this technology-based arena, brain tumor images with proper labeling and acquisition still have a problem with the accurate and reliable generation of realistic images of brain tumors that are completely different from the original ones. The artificially created medical image data would help improve the learning ability of physicians and other computer-aided systems for the generation of augmented data. To overcome the highlighted issue, a Generative Adversarial Network (GAN) deep learning technique in which two neural networks compete to become more accurate in creating artificially realistic data for MRI images. The GAN network contains mainly two parts known as generator and discriminator. Commonly, a generator is the convolutional neural network, and a discriminator is the deconvolutional neural network. In this research, the publicly accessible Contrast-Enhanced Magnetic Resonance Imaging (CE-MRI) dataset collected from 2005-to 2020 from different hospitals in China consists of four classes has been used. Our proposed method is simple and achieved an accuracy of 96%. We compare our technique results with the existing results, indicating that our proposed technique outperforms the best results associated with the existing methods.

Machine Learning for Surrogate Groundwater Modelling of a Small Carbonate Island

Barbados is heavily reliant on groundwater resources for its potable water supply, with over 80% of the island’s water sourced from aquifers. The ability to meet demand will become even more challenging due to the continuing climate crisis. The consequences of climate change within the Caribbean region include sea level rise, as well as hydrometeorological effects such as increased rainfall intensity, and declines in average annual rainfall. Scientifically sound approaches are becoming increasingly important to understand projected changes in supply and demand while concurrently minimizing deleterious impacts on the island’s aquifers. Therefore, the objective of this paper is to develop a physics-based groundwater model and surrogate models using machine learning (ML), which provide decision support to assist with groundwater resources management in Barbados. Results from the study show that a single continuum conceptualization is adequate for representing the island’s hydrogeology as demonstrated by a root mean squared error and mean absolute error of 2.7 m and 2.08 m between the model and observed steady-state hydraulic head. In addition, we show that data-driven surrogates using deep neural networks, elastic networks, and generative adversarial networks are capable of approximating the physics-based model with a high degree of accuracy as shown by R-squared values of 0.96, 0.95, and 0.95, respectively. The framework and tools developed are a critical step towards a digital twin that provides stakeholders with a quantitative tool for optimal management of groundwater under a changing climate in Barbados. These outputs will provide sound evidence-based solutions to aid long-term economic and social development on the island.

Developing a data-driven technology roadmapping method using generative adversarial network (GAN)

The technology roadmap is used as an essential tool to establish strategies. Existing expert-based and data-based roadmapping processes consume a lot of resources, are difficult to update, and can be biased towards experts’ subjectivity. To overcome these limitations, this study proposes an automated deep learning model Generative Adversarial Network (GAN) based technology roadmapping method. The proposed framework consists of two modules. Module 1 uses the GAN model to train the Roadmap, and trains the GAN model by the knowledge of experts in the technology roadmap using the existing technology roadmap data. The defined model consists of a generator that receives the node of the technology roadmap and generates the next node, and discriminators that verify the generated node, and provide feedback. Module 2 expands the roadmap, generates candidate nodes using the trained GAN model, and merges suitable nodes into the roadmap. To select the appropriate configuration for the proposed framework, experiments were conducted in terms of data format and model layer. Also, to validate the model presented in this study, the analysis was performed by selecting the renewable energy industry as a target technology of the analysis.

Deep learning to harmonize MRI scans for better diagnosis and prognosis in multi‐center studies

AbstractBackgroundData heterogeneity due to cross‐scanner variations is a major challenge in multi‐center aging and dementia studies. Brain structural measures for the same participant can vary due to scanner differences (manufacturer, scanner age/technology, signal‐to‐noise ratio, pulse sequence design, contrast, resolution) and result in biomarker variability. In this work, we applied basic image processing and deep learning methods (style transfer (ST) and generative adversarial networks (GANs)) for harmonizing T1‐weighted MRI scans across change in vendors (GE to Siemens) and resolution (1x1x1.2mm to 0.8x0.8x0.8mm) and test whether this harmonization improved compatibility of AD biomarker measures.MethodWe utilized training (n=814) and test (n=113 same participants scanned on two scanners close‐in‐time, typically on the same day, with different resolution) data from cognitively impaired and unimpaired participants scanned on GE and Siemens scanners from Mayo Clinic Study of Aging and Mayo ADRC. The harmonization was conducted to translate GE to Siemens. All scans were affine transformed to the same template space, resampled to 1.5 mm3, and bias corrected. Histogram matching was performed using cumulative density functions. A ResNet18 was trained for the ST’s content and style layers. Two types of GANs, conditional GANs (cGAN) and cycle‐consistent GANs (cycleGAN), were constructed. We then measured regional cortical thickness from each image using FreeSurfer‐7.1.1 and compared the thickness measurements from the original and translated scans using intraclass‐correlation (ICC) and Deming regression.ResultIn the test dataset, ST and cycleGAN showed the best performances when comparing the translated GE to Siemens. With ST, the ICC increased (vs. the original scans) slightly for the temporal meta‐ROI (0.956 from 0.954) and more substantially for the frontal pole ROI (0.706 from 0.594). The cycleGAN showed an increase in ICC for the frontal pole ROI (0.610 from 0.594) but not the meta‐ROI (Fig. 1).ConclusionEach method leveraged a different strength. Histogram matching captures contrast, ST extracts textural information, and GANs learn combination of contrast, edge and texture properties. We found that deep learning is a promising approach to cross‐scanner variability. Our comparisons revealed anatomical differences across methods; thus, disease specific measure should be considered for selection of the scanner harmonization method.

Automated urban planning aware spatial hierarchies and human instructions

Traditional urban planning demands urban experts to spend much time producing an optimal urban plan under many architectural constraints. The remarkable imaginative ability of deep generative learning provides hope for renovating this domain. Existing works are constrained by: (1) neglecting human requirements; (2) omitting spatial hierarchies, and (3) lacking urban plan samples. We propose a novel, deep human-instructed urban planner to fill these gaps and implement two practical frameworks. In the preliminary version, we formulate the task into an encoder–decoder paradigm. The encoder is to learn the information distribution of surrounding contexts, human instructions, and land-use configuration. The decoder is to reconstruct the land-use configuration and the associated urban functional zones. Although it has achieved good results, the generation performance is still unstable due to the complex optimization directions of the decoder. Thus, we propose a cascading deep generative adversarial network (GAN) in this paper, inspired by the workflow of urban experts. The first GAN is to build urban functional zones based on human instructions and surrounding contexts. The second GAN will produce the land-use configuration by considering the built urban functional zones. Finally, we conducted extensive experiments and case studies to validate the effectiveness and superiority of our work.

Relaxed deep generative adversarial networks for real-time economic smart generation dispatch and control of integrated energy systems

The randomness of renewable energy connected to power systems cannot satisfy the frequency stability and economic dispatch of integrated energy systems. The conventional combined dispatch and control framework can benefit the interests of only a part of energy systems, such as gas/heat/cold systems. Real-time economic smart generation dispatch and control are applied to solve these two problems of integrated energy systems. This paper proposes a relaxed deep generative adversarial network method for the real-time control framework to replace conventional combined methods with multiple time-scale combined frameworks. The proposed approach combines deep learning and relaxed operation with generative adversarial networks as the proposed method. Because of the powerful representation capability of deep learning and the advantages of amplitude constraints of the relaxation operation, the proposed method can simultaneously generate multiple convergent and high-performance generation commands. The proposed method is simulated for simple and large-scale complex integrated energy systems based on a 14-bus with ten natural-gas systems. Furthermore, the generation controller based on the proposed method can economically satisfy the frequency stability for integrated energy systems, mitigate uneven energy distribution and excess energy, improve energy efficiency, and achieve multi-energy complementarity.

Diagnosing Brain Tumors Using a Super Resolution Generative Adversarial Network Model

Аutоmаted deteсtiоn оf tumоrs in MRIs is inсredibly vital as it рrоvides details аbоut аnomalous tissues that are imроrtаnt fоr рlаnning further pathways of treаtment. It is an imрrасtiсаl method requiring massive аmоunt оf knоwledge. Henсe, trustworthy аnd аutоmаtiс сlаssifiсаtiоn sсhemes and рrоgrаmmes аre сruсiаl to put an end to the deаth rаte оf humаns. Sо, deteсtiоn methods аre developed that wоuld not only save the time of the radiologist but also help in асquiring а tested ассurасy. Manual detection of MRI tumor соuld be а соmрliсаted tаsk due tо the соmрlexity аnd vаriаnсe оf tumоrs. In this paper, the authors рrороse both mасhine leаrning and deep learning-based generative adversarial network (GAN) аlgоrithms tо overcome the challenges оf conventional сlаssifiers where tumоrs were deteсted in brаin MRIs using mасhine leаrning аlgоrithms only. Making use of SR-GAN increases the accuracy of the proposed method to more than 98%.

Visual Saliency and Image Reconstruction from EEG Signals via an Effective Geometric Deep Network-Based Generative Adversarial Network

Reaching out the function of the brain in perceiving input data from the outside world is one of the great targets of neuroscience. Neural decoding helps us to model the connection between brain activities and the visual stimulation. The reconstruction of images from brain activity can be achieved through this modelling. Recent studies have shown that brain activity is impressed by visual saliency, the important parts of an image stimuli. In this paper, a deep model is proposed to reconstruct the image stimuli from electroencephalogram (EEG) recordings via visual saliency. To this end, the proposed geometric deep network-based generative adversarial network (GDN-GAN) is trained to map the EEG signals to the visual saliency maps corresponding to each image. The first part of the proposed GDN-GAN consists of Chebyshev graph convolutional layers. The input of the GDN part of the proposed network is the functional connectivity-based graph representation of the EEG channels. The output of the GDN is imposed to the GAN part of the proposed network to reconstruct the image saliency. The proposed GDN-GAN is trained using the Google Colaboratory Pro platform. The saliency metrics validate the viability and efficiency of the proposed saliency reconstruction network. The weights of the trained network are used as initial weights to reconstruct the grayscale image stimuli. The proposed network realizes the image reconstruction from EEG signals.

Probabilistic Wind Park Power Prediction using Bayesian Deep Learning and Generative Adversarial Networks

The rapid depletion of fossil-based energy supplies, along with the growing reliance on renewable resources, has placed supreme importance on the predictability of renewables. Research focusing on wind park power modelling has mainly been concerned with point estimators, while most probabilistic studies have been reserved for forecasting. In this paper, a few different approaches to estimate probability distributions for individual turbine powers in a real off-shore wind farm were studied. Two variational Bayesian inference models were used, one employing a multilayered perceptron and another a graph neural network (GNN) architecture. Furthermore, generative adversarial networks (GAN) have recently been proposed as Bayesian models and was here investigated as a novel area of research. The results showed that the two Bayesian models outperformed the GAN model with regards to mean absolute errors (MAE), with the GNN architecture yielding the best results. The GAN on the other hand, seemed potentially better at generating diverse distributions. Standard deviations of the predicted distributions were found to have a positive correlation with MAEs, indicating that the models could correctly provide estimates on the confidence associated with particular predictions.

Image synthesis of effective atomic number images using a deep convolutional neural network-based generative adversarial network.

The effective atomic numbers obtained from dual-energy computed tomography (DECT) can aid in characterization of materials. In this study, an effective atomic number image reconstructed from a DECT image was synthesized using an equivalent single-energy CT image with a deep convolutional neural network (CNN)-based generative adversarial network (GAN). The image synthesis framework to obtain the effective atomic number images from a single-energy CT image at 120 kVp using a CNN-based GAN was developed. The evaluation metrics were the mean absolute error (MAE), relative root mean square error (RMSE), relative mean square error (MSE), structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), and mutual information (MI). The difference between the reference and synthetic effective atomic numbers was within 9.7% in all regions of interest. The averages of MAE, RMSE, MSE, SSIM, PSNR, and MI of the reference and synthesized images in the test data were 0.09, 0.045, 0.0, 0.89, 54.97, and 1.03, respectively. In this study, an image synthesis framework using single-energy CT images was constructed to obtain atomic number images scanned by DECT. This image synthesis framework can aid in material decomposition without extra scans in DECT.

Generating Synthetic Images for Healthcare with Novel Deep Pix2Pix GAN

Due to recent developments in deep learning and artificial intelligence, the healthcare industry is currently going through a significant upheaval. Despite a considerable advance in medical imaging and diagnostics, the healthcare industry still has a lot of unresolved problems and unexplored applications. The transmission of a huge number of medical images in particular is a difficult and time-consuming problem. In addition, obtaining new medical images is too expensive. To tackle these issues, we propose deep pix2pix generative adversarial networks (GAN) for generating synthetic medical images. For the comparison, we implemented CycleGAN, Pix2Pix GAN and Deep Pix2Pix GAN. The result has shown that our proposed approach can generate a new synthetic medical image from a different image with more accuracy than that of the other models. To provide a robust model, we trained and evaluated our models on a widely used brain image dataset, the IXI Dataset.

Modified GAN for Natural Occlusion Detection and Inpainting of Raw Footage From Video Surveillance

Low resolution and occlusion are mainly prominent in images taken from certain unconstrained environments such as raw footage from video surveillance. In this work, a deep generative adversarial network for joint face completion and face super-resolution is proposed. It will be really useful in the current COVID-19 scenario as people wearing masks are a common sight. Given an input of a low-resolution face image with occlusion, the generator aims to recover a high-resolution face image without occlusion. The discriminator uses a set of carefully designed losses to assure the high quality of the recovered high-resolution face images without occlusion. Experimental results on CelebA database show that the proposed approach outperforms the state-of-the-art methods in jointly performing face super-resolution and face completion, and shows good generalization ability in cross-database testing. MSSIM showed an accuracy of around 80% for test cases. The recorded values of generator adversarial loss, generator pixel loss, and discriminator loss are 0.93, 0.10, and 0.003, respectively.