Conditional Generative Adversarial Network Model Research Articles

Damage detection for railway bridges using time‐frequency decomposition and conditional generative model

AbstractA novel damage detection model, which utilizes the spatiotemporal characteristics of the acceleration data, is proposed to assess the structural integrity of railway bridges. For this, the measured acceleration data are decomposed into several intrinsic mode functions (IMFs) using the sparse random mode decomposition model. The generated IMFs are subsequently integrated into the enhanced time series conditional generative adversarial network model to identify possible damage in bridges across various frequency bands. The influence of environmental and operational variables (EOVs), particularly temperature fluctuations, was also investigated. The proposed model was verified using both numerical and experimental data from a plate girder bridge. Further validation was conducted using the Z24 bridge dataset, and damage cases under the influence of EOVs were successfully predicted. Throughout the validation process, various anomaly metrics were introduced to establish a threshold value, and a covariance‐based time domain metric was proven to be the most effective in our cases.

Airfoil aerodynamic/stealth design based on conditional generative adversarial networks

Aerodynamic/stealth design is becoming an important factor in the advanced airfoil design. In this work, a supervised machine learning method is proposed for aerodynamic and stealth integrated airfoil design. The conditional generative adversarial network (CGAN) is constructed for the multidisciplinary design of airfoil. Then, the generator and discriminator simply using deep neural network have good robustness and stability in training. The CGAN model also shows good generalization capability in the test set, with less than 1% error in fitting to the airfoil profile data, and the generated airfoils are within 10% error compared to the test airfoil aerodynamic stealth characteristics. In addition, the optimization results based on the CGAN model demonstrate that aerodynamic performance improvement would increase the airfoil camber and stealth performance improvement would sharpen the airfoil leading edge.

Generation of Vessel Track Characteristics Using a Conditional Generative Adversarial Network (CGAN)

ABSTRACT Machine learning (ML) models often require large volumes of data to learn a given task. However, access and existence of training data can be difficult to acquire due to privacy laws and availability. A solution is to generate synthetic data that represents the real data. In the maritime environment, the ability to generate realistic vessel positional data is important for the development of ML models in ocean areas with scarce amounts of data, such as the Arctic, or for generating an abundance of anomalous or unique events needed for training detection models. This research explores the use of conditional generative adversarial networks (CGAN) to generate vessel displacement tracks over a 24-hour period in a constraint-free environment. The model is trained using Automatic Identification System (AIS) data that contains vessel tracking information. The results show that the CGAN is able to generate vessel displacement tracks for two different vessel types, cargo ships and pleasure crafts, for three months of the year (May, July, and September). To evaluate the usability of the generated data and robustness of the CGAN model, three ML vessel classification models using displacement track data are developed using generated data and tested with real data.

Prediction for underground seismic intensity measures using conditional generative adversarial networks

With the escalating development and utilization of subterranean spaces, the seismic hazards faced by underground structures are progressively increasing. However, owing to the challenges associated with acquiring underground seismic data and historical seismic design norms, pertinent regulations and research in this domain are scarce. This study focused on three crucial intensity measures in the seismic design process of underground structures peak ground acceleration (PGA), peak ground velocity (PGV), and peak ground displacement (PGD). The research leverages seismic data obtained from the California Strong Motion Instrumentation Program (CSMIP) to train and evaluate a conditional generative adversarial network (CGAN) model. This model was employed to establish a multivariate joint conditional probability distribution among the intensity measures at varying depths, facilitating the stochastic prediction of shallow intensity measures. In contrast to empirical formulas, the CGAN model eliminates the need for a predefined equation structure and enables the simultaneous prediction of multiple intensity measures. The performance of the model was evaluated by comparing the predictive accuracy of the CGAN model and empirical fitting formulas across diverse site conditions and depth intervals using metrics such as relative error coefficients. It can be concluded that the proposed CGAN model can accurately predict shallow seismic intensity measures, and the predictions conform to a specific conditional distribution while retaining the stochastic nature of seismic motion. Compared with empirical formula models, the CGAN model exhibited an enhanced predictive capability.

Data-driven framework for prediction and optimization of gas turbine blade film cooling

Film cooling is a crucial technique for protecting critical components of gas turbines from excessive temperatures. Multiparameter film cooling optimization is still relatively time-consuming owing to the substantial computational demands of computational fluid dynamics (CFD) methods. To reduce the computational cost, the present study develops a data-driven framework for predicting and optimizing the film cooling effectiveness of high-pressure turbines based on deep learning. Multiple rows of cooling holes located on the pressure surface of the turbine blade are optimized, with the coolant hole diameter, the incline angle, and the compound angle as design parameters. A conditional generative adversarial network model combining a gated recurrent unit and a convolutional neural network is designed to establish the complex nonlinear regression between the design parameters and the film cooling effectiveness. The surrogate model is trained and tested using independent CFD results. A sparrow search algorithm and the well-trained surrogate model are combined to acquire the optimal film cooling parameters. The proposed framework is found to improve multi-row film cooling effectiveness by 21.2% at an acceptable computational cost.

CGAN-ECT: Reconstruction of Electrical Capacitance Tomography images from capacitance measurements using Conditional Generative Adversarial Networks

Due to the rapid growth of applications of Electrical Capacitance Tomography (ECT) across diverse industrial sectors, there is a pressing demand for swift and high-fidelity methodologies for image reconstruction from raw capacitance measurements. Deep learning, as an effective non-linear mapping tool for complicated functions, has been going viral in many fields including electrical tomography. This paper introduces a pioneering contribution by proposing a Conditional Generative Adversarial Network (CGAN) model explicitly tailored for the purpose of reconstructing ECT images from capacitance measurements. We propose a novel modulator to transform the raw input capacitance measurements into a 2D image. To the best of our knowledge, this is the first time to represent the capacitance measurements in an image form. We have created a new massive ECT dataset of synthetic image-measurements pairs for training and testing purposes of the proposed model. This extensive ECT dataset comprises 320,000 synthetic image-measurement pairs. It forms the basis for evaluating the feasibility and generalizability of the CGAN-ECT model, with assessments conducted on testing data, contaminated data, and flow patterns which are intentionally excluded from the model’s training phase. The evaluation results prove the efficacy of the proposed CGAN model in generating ECT images with heightened accuracy compared to traditional methods and alternative deep learning-based image reconstruction algorithms. The CGAN-ECT model has succeeded in a achieving an average image correlation coefficient exceeding 99.3%, coupled with an average relative image error of approximately 0.07.

Artificial intelligence empowering museum space layout design: Insights from China.

The floor plan layout of museum exhibition spaces is the skeleton network of the museum, which determines the internal circulation and spatial form of the museum. This paper studies the method and practice of using artificial intelligence technology to assist in the space design of exhibition halls in urban cultural museums. First, it introduces the limitations of traditional space design methods for exhibition halls in urban cultural museums and the superiority and application prospects of the CGAN (conditional generative adversarial network) model in space design. Second, the principle and training process of the CGAN model are explained in detail, and the experimental results and analysis are given. By learning 100 floor plans of exhibition halls of urban culture museums, the CGAN model can generate a new floor plan design for an exhibition hall, which provides a new idea and innovative method for this design task. Finally, the limitations and future research directions of the CGAN model in the space design of urban cultural museum exhibition halls are discussed. The study shows that using the CGAN model to learn the floor plans of exhibition halls of urban cultural museums can effectively improve the innovation and practicability of space design and has the following advantages: (1) It can quickly generate a large number of exhibition hall floor plans, shorten the design cycle, and improve design efficiency. (2) The generated floor plan designs of the exhibition hall are diverse and personalized, meeting the design requirements of different scenarios and needs. (3) The method promotes the deep integration of space design and artificial intelligence technology and provides new possibilities and ideas for space design. These conclusions provide new ideas and methods for the space design of exhibition halls of urban cultural museums and provide a reference and inspiration for space design and intelligent applications in other fields, such as office space design, home decoration space design, landscape space design, and historical arcade and building renovation design.

Next-Gen brain tumor classification: pioneering with deep learning and fine-tuned conditional generative adversarial networks

Brain tumor has become one of the fatal causes of death worldwide in recent years, affecting many individuals annually and resulting in loss of lives. Brain tumors are characterized by the abnormal or irregular growth of brain tissues that can spread to nearby tissues and eventually throughout the brain. Although several traditional machine learning and deep learning techniques have been developed for detecting and classifying brain tumors, they do not always provide an accurate and timely diagnosis. This study proposes a conditional generative adversarial network (CGAN) that leverages the fine-tuning of a convolutional neural network (CNN) to achieve more precise detection of brain tumors. The CGAN comprises two parts, a generator and a discriminator, whose outputs are used as inputs for fine-tuning the CNN model. The publicly available dataset of brain tumor MRI images on Kaggle was used to conduct experiments for Datasets 1 and 2. Statistical values such as precision, specificity, sensitivity, F1-score, and accuracy were used to evaluate the results. Compared to existing techniques, our proposed CGAN model achieved an accuracy value of 0.93 for Dataset 1 and 0.97 for Dataset 2.

Spatial extrapolation of downscaled geochemical data using conditional GAN

Multi-scale geochemical data play an important guiding role in mineral exploration and environmental assessment. However, due to the influence of certain factors such as terrain, sampling cost and time, and instrument detection limit, geochemical data obtained from a study area are often incomplete or sparse. In this study, a conditional generative adversarial network (GAN) model was developed as a stochastic simulation technique for spatial extrapolation of downscaled geochemical data by learning nonlinear relationships between two soil geochemical datasets with different sampling densities, one set called fine data and the other coarse data. Taking Qianjiang City, Hubei Province, China as an example, first the fine data and coarse data were divided separately into training area and verification area for simulation in the GAN model. Then, exploratory data analysis, error analysis and variogram analysis were used to evaluate the realizations. Finally, Ni anomalies related to soil heavy metal pollution were extracted by combined concentration–area (C–A) fractal model and uncertainty analysis, in the verification area. Our studies indicate that the proposed conditional GAN model is efficient in extrapolating downscaled geochemical data, since both the realizations and fine data maintain similar spatial autocorrelation and statistical distribution.

Smart City Facility Location Recommendation Algorithm Using Multimedia Data and Improved Generative Adversarial Networks

Internet of Things, low energy consumption, and intelligent and functionally integrated urban infrastructure construction are crucial elements in the development of smart cities. Distributed generation (DG) and electric vehicle charging infrastructure play a vital role in the planning and construction of smart cities. However, the uncertainty associated with the power output of distributed generation significantly impacts the planning of distribution networks. To address this issue, this paper proposes a site-selection recommendation algorithm that leverages urban multimedia data and improved generative adversarial networks. The proposed algorithm begins by modeling the uncertainty of wind power and photovoltaic (PV) generation using an enhanced conditional generative adversarial network model. To generate multimedia datasets with time-series characteristics for wind power and PV generation scenarios, monthly multimedia data labels are incorporated into the model. These multimedia datasets, representing a wide range of scenarios, are then clustered using the K-means clustering method. Furthermore, a distributed generation planning model is established, aiming to minimize the annual integrated cost. The planning problem is efficiently solved using CPLEX, a mathematical programming solver. In the simulation experiments, the proposed scheme is compared with alternative schemes. The results demonstrate that the proposed scheme achieves a significant total cost saving of 21.95% compared to the comparison scheme. Moreover, the experimental comparison reveals that the proposed scheme exhibits higher stability. Additionally, in terms of algorithm efficiency, the proposed algorithm outperforms the other three algorithms tested in terms of the number of iterations and speed. The experimental results highlight the effectiveness of the proposed planning model in improving the economy and stability of the distribution network. Furthermore, it enhances the computational efficiency of the planning problem associated with distributed power supply and electric vehicle charging stations. The findings of this research hold substantial research significance for the site selection planning of distributed power supply.

Generating synthetic personal health data using conditional generative adversarial networks combining with differential privacy

A large amount of personal health data that is highly valuable to the scientific community is still not accessible or requires a lengthy request process due to privacy concerns and legal restrictions. As a solution, synthetic data has been studied and proposed to be a promising alternative to this issue. However, generating realistic and privacy-preserving synthetic personal health data retains challenges such as simulating the characteristics of the patients’ data that are in the minority classes, capturing the relations among variables in imbalanced data and transferring them to the synthetic data, and preserving individual patients’ privacy. In this paper, we propose a differentially private conditional Generative Adversarial Network model (DP-CGANS) consisting of data transformation, sampling, conditioning, and network training to generate realistic and privacy-preserving personal data. Our model distinguishes categorical and continuous variables and transforms them into latent space separately for better training performance. We tackle the unique challenges of generating synthetic patient data due to the special data characteristics of personal health data. For example, patients with a certain disease are typically the minority in the dataset and the relations among variables are crucial to be observed. Our model is structured with a conditional vector as an additional input to present the minority class in the imbalanced data and maximally capture the dependency between variables. Moreover, we inject statistical noise into the gradients in the networking training process of DP-CGANS to provide a differential privacy guarantee. We extensively evaluate our model with state-of-the-art generative models on personal socio-economic datasets and real-world personal health datasets in terms of statistical similarity, machine learning performance, and privacy measurement. We demonstrate that our model outperforms other comparable models, especially in capturing the dependence between variables. Finally, we present the balance between data utility and privacy in synthetic data generation considering the different data structures and characteristics of real-world personal health data such as imbalanced classes, abnormal distributions, and data sparsity.

Evaluation of motion artefact reduction depending on the artefacts’ directions in head MRI using conditional generative adversarial networks

Motion artefacts caused by the patient’s body movements affect magnetic resonance imaging (MRI) accuracy. This study aimed to compare and evaluate the accuracy of motion artefacts correction using a conditional generative adversarial network (CGAN) with an autoencoder and U-net models. The training dataset consisted of motion artefacts generated through simulations. Motion artefacts occur in the phase encoding direction, which is set to either the horizontal or vertical direction of the image. To create T2-weighted axial images with simulated motion artefacts, 5500 head images were used in each direction. Of these data, 90% were used for training, while the remainder were used for the evaluation of image quality. Moreover, the validation data used in the model training consisted of 10% of the training dataset. The training data were divided into horizontal and vertical directions of motion artefact appearance, and the effect of combining this data with the training dataset was verified. The resulting corrected images were evaluated using structural image similarity (SSIM) and peak signal-to-noise ratio (PSNR), and the metrics were compared with the images without motion artefacts. The best improvements in the SSIM and PSNR were observed in the consistent condition in the direction of the occurrence of motion artefacts in the training and evaluation datasets. However, SSIM > 0.9 and PSNR > 29 dB were accomplished for the learning model with both image directions. The latter model exhibited the highest robustness for actual patient motion in head MRI images. Moreover, the image quality of the corrected image with the CGAN was the closest to that of the original image, while the improvement rates for SSIM and PSNR were approximately 26% and 7.7%, respectively. The CGAN model demonstrated a high image reproducibility, and the most significant model was the consistent condition of the learning model and the direction of the appearance of motion artefacts.

Application of conditional generative adversarial network to multi-step car-following modeling.

Car-following modeling is essential in the longitudinal control for connected and autonomous vehicles (CAVs). Considering the advantage of the generative adversarial network (GAN) in capturing realistic data distribution, this paper applies conditional GAN (CGAN) to car-following modeling. The generator is elaborately designed with a sequence-to-sequence structure to reflect the decision-making process of human driving behavior. The proposed model is trained and tested based on the empirical dataset, and it is compared with a supervised learning model and a mathematical model. Numerical simulations are conducted to verify the model's performance, especially in the condition of mixed traffic flow. The comparison result shows that the CGAN model outperforms others in trajectory reproduction, indicating it can effectively imitate human driving behavior. The simulation results suggest that the introduction of CGAN-based CAVs improves the stability and efficiency of the mixed traffic flow.

Real-World Underwater Image Enhancement Based on Attention U-Net

In recent years, with the increasingly serious problems of resource shortage and environmental pollution, the exploration and development of underwater clean energy were particularly important. At the same time, abundant underwater resources and species have attracted a large number of scientists to carry out research on underwater-related tasks. Due to the diversity and complexity of underwater environments, it is difficult to perform related vision tasks, such as underwater target detection and capture. The development of digital image technology has been relatively mature, and it has been applied in many fields and achieved remarkable results, but the research on underwater image processing technology is rarely effective. The underwater environment is much more complicated than that on land, and there is no light source underwater. Underwater imaging systems must rely on artificial light sources for illumination. When light travels through water, it is severely attenuated by water absorption, reflection, and scattering. The collected underwater images inevitably have problems such as limited visible range, blur, low contrast, uneven illumination, incoherent colors, and noise. The purpose of image enhancement is to improve or solve one or more of the above problems in a targeted manner. Therefore, underwater image enhancement technology has become one of the key contents of underwater image processing technology research. In this paper, we proposed a conditional generative adversarial network model based on attention U-Net which contains an attention gate mechanism that could filter invalid feature information and capture contour, local texture, and style information effectively. Furthermore, we formulate an objective function through three different loss functions, which can evaluate image quality from global content, color, and structural information. Finally, we performed end-to-end training on the UIEB real-world underwater image dataset. The comparison experiments show that our method outperforms all comparative methods, the ablation experiments show that the loss function proposed in this paper outperforms a single loss function, and finally, the generalizability of our method is verified by executing on two different datasets, UIEB and EUVP.

The Floor Plan Design Method of Exhibition Halls in CGAN-Assisted Museum Architecture

The floor plan designs of traditional museum exhibition halls are generally developed according to the position and streamlined accessibility of the exhibits. However, there are often many floors in the same building, and multi-story exhibition halls are similar, so architects often spend a large amount of time and energy designing floors individually. Thus, this paper proposes a conditional generative adversarial network (CGAN)-based method for designing the floor plans of museum exhibition halls, which can help architects to work more efficiently. In this study, the basic concepts and structures of CGAN are first introduced; then, the design and training process of the CGAN model used are described in detail, and the datasets and evaluation metrics adopted are briefly described. In the Results and Discussion sections, this paper presents an example of the generated floor plan design of a museum exhibition hall and evaluates and analyzes the floor plan design of a museum exhibition hall generated using the proposed method. Finally, this paper summarizes the advantages of the proposed method, but also notes its shortcomings. If the number of data sets is not sufficient, the scope of the application will be relatively small. For example, museums converted from certain historical buildings are not applicable. The research results show the following: (1) the method proposed in this paper takes advantage of the CGAN model and can generate a museum exhibition hall floor plan design with certain regularity according to the given conditions, rather than pure random generation. (2) This method can automatically generate a variety of plan designs for museum exhibition halls in different schemes, providing designers with more choices and flexibility. (3) This method can carry out design optimization through human–computer interaction, and iterative improvement can be carried out according to user needs and feedback, which improves the practicability of the design.

Fast prediction of indoor airflow distribution inspired by synthetic image generation artificial intelligence.

Prediction of indoor airflow distribution often relies on high-fidelity, computationally intensive computational fluid dynamics (CFD) simulations. Artificial intelligence (AI) models trained by CFD data can be used for fast and accurate prediction of indoor airflow, but current methods have limitations, such as only predicting limited outputs rather than the entire flow field. Furthermore, conventional AI models are not always designed to predict different outputs based on a continuous input range, and instead make predictions for one or a few discrete inputs. This work addresses these gaps using a conditional generative adversarial network (CGAN) model approach, which is inspired by current state-of-the-art AI for synthetic image generation. We create a new Boundary Condition CGAN (BC-CGAN) model by extending the original CGAN model to generate 2D airflow distribution images based on a continuous input parameter, such as a boundary condition. Additionally, we design a novel feature-driven algorithm to strategically generate training data, with the goal of minimizing the amount of computationally expensive data while ensuring training quality of the AI model. The BC-CGAN model is evaluated for two benchmark airflow cases: an isothermal lid-driven cavity flow and a non-isothermal mixed convection flow with a heated box. We also investigate the performance of the BC-CGAN models when training is stopped based on different levels of validation error criteria. The results show that the trained BC-CGAN model can predict the 2D distribution of velocity and temperature with less than 5% relative error and up to about 75,000 times faster when compared to reference CFD simulations. The proposed feature-driven algorithm shows potential for reducing the amount of data and epochs required to train the AI models while maintaining prediction accuracy, particularly when the flow changes non-linearly with respect to an input.

Conditional Generative Adversarial Network for Monocular Image Depth Map Prediction

Deep map prediction plays a crucial role in comprehending the three-dimensional structure of a scene, which is essential for enabling mobile robots to navigate autonomously and avoid obstacles in complex environments. However, most existing depth estimation algorithms based on deep neural networks rely heavily on specific datasets, resulting in poor resistance to model interference. To address this issue, this paper proposes and implements an optimized monocular image depth estimation algorithm based on conditional generative adversarial networks. The goal is to overcome the limitations of insufficient training data diversity and overly blurred depth estimation contours in current monocular image depth estimation algorithms based on generative adversarial networks. The proposed algorithm employs an enhanced conditional generative adversarial network model with a generator that adopts a network structure similar to UNet and a novel feature upsampling module. The discriminator uses a multi-layer patchGAN conditional discriminator and incorporates the original depth map as input to effectively utilize prior knowledge. The loss function combines the least squares loss function and the L1 loss function. Compared to traditional depth estimation algorithms, the proposed optimization algorithm can effectively restore image contour information and enhance the visualization capability of depth prediction maps. Experimental results demonstrate that our method can expedite the convergence of the model on NYU-V2 and Make3D datasets, and generate predicted depth maps that contain more details and clearer object contours.

Conditional Generative Adversarial Network Model for Conversion of 2 Dimensional Radiographs Into 3 Dimensional Views

Conditional Generative Adversarial Network Model for Conversion of 2 Dimensional Radiographs Into 3 Dimensional Views

Deep Neural Network-Based Generation of Planar CH Distribution through Flame Chemiluminescence in Premixed Turbulent Flame

Flame front structure is one of the most fundamental characteristics and, hence, vital for understanding combustion processes. Measuring flame front structure in turbulent flames usually needs laser-based diagnostic techniques, mostly planar laser-induced fluorescence (PLIF). The equipment of PLIF, burdened with lasers, is often too sophisticated to be configured in harsh environments. Here, to shed the burden, we propose a deep neural network-based method to generate the structures of flame fronts using line-of-sight CH* chemiluminescence that can be obtained without the use of lasers. A conditional generative adversarial network (C-GAN) was trained by simultaneously recording CH-PLIF and chemiluminescence images of turbulent premixed methane/air flames. Two distinct generators of the C-GAN, namely Resnet and U-net, were evaluated. The former net performs better in this study in terms of both generating snap-shot images and statistics over multiple images. For chemiluminescence imaging, the selection of the camera's gate width produces a trade-off between the signal-to-noise (SNR) ratio and the temporal resolution. The trained C-GAN model can generate CH-PLIF images from the chemiluminescence images with an accuracy of over 91% at a Reynolds number of 5000, and the flame surface density at a higher Reynolds number of 10,000 can also be effectively estimated by the model. This new method has the potential to achieve the flame characteristics without the use of laser and significantly simplify the diagnosing system, also with the potential for high-speed flame diagnostics.

Deep neural network for oil spill detection using Sentinel-1 data: application to Egyptian coastal regions

Building an oil spill segmentation model is very challenging because of the limited available information on oil spill accidents. Therefore, this paper proposes a custom data generator based on Segmentation Network (Seg-Net) model implemented in Conditional Generative Adversarial Network (CGAN). The proposed model is trained for oil spill segmentation using 50 Sentinal-1 Synthetic Aperture Radar (SAR) images. The proposed model employs a modified Seg-Net as a generator to produce high-quality oil spills’ images and a Patch-GAN as discriminator. This architecture results in a significant improvement of the final oil segmentation results, in comparison with Seg-Net model, while using relatively small training dataset. For performance assessment, the paper presents the oil spills segmentation results of four suggested models using Sentinel-1 SAR images. The presented models are U-Net, Seg-Net, CGAN, and a Seg-Net-based CGAN the performance assessment reveals that the proposed model produces oil spill segmentation images with an average accuracy of 99.04%, Intersection over Union (IoU) index of 96.59%, and a precision of 85.24%. In addition, the training time required for the proposed model is 3 h 20 min per 50 epochs, while it is nearly 10 h and 55 min for training a CGAN model.