scholarly journals GANTL: Towards Practical and Real-Time Topology Optimization with Conditional GANs and Transfer Learning

2021 ◽  
pp. 1-32
Author(s):  
Mohammad Mahdi Behzadi ◽  
Horea T. Ilies
Keyword(s):  
Deep Learning ◽  
Topology Optimization ◽  
Knowledge Transfer ◽  
Boundary Conditions ◽  
Transfer Learning ◽  
Computational Cost ◽  
Generative Adversarial Networks ◽  
Learning Methods ◽  
Adversarial Networks ◽  
Difficult Time

Abstract Many machine learning methods have been recently developed to circumvent the high computational cost of the gradient-based topology optimization. These methods typically require extensive and costly datasets for training, have a difficult time generalizing to unseen boundary and loading conditions and to new domains, and do not take into consideration topological constraints of the predictions, which produces predictions with inconsistent topologies. We present a deep learning method based on generative adversarial networks for generative design exploration. The proposed method combines the generative power of conditional GANs with the knowledge transfer capabilities of transfer learning methods to predict optimal topologies for unseen boundary conditions. We also show that the knowledge transfer capabilities embedded in the design of the proposed algorithm significantly reduces the size of the training dataset compared to the traditional deep learning neural or adversarial networks. Moreover, we formulate a topological loss function based on the bottleneck distance obtained from the persistent diagram of the structures and demonstrate a significant improvement in the topological connectivity of the predicted structures. We use numerous examples to explore the efficiency and accuracy of the proposed approach for both seen and unseen boundary conditions in 2D.

scholarly journals GGADN: Guided Generative Adversarial Dehazing Network

2021 ◽  
Author(s):  
Zhang Jian ◽  
Wanjuan Song
Keyword(s):  
Deep Learning ◽  
Loss Function ◽  
Minimax Problem ◽  
Feature Model ◽  
Generative Adversarial Networks ◽  
Image Dehazing ◽  
Learning Methods ◽  
Adversarial Networks ◽  
Network Training

Abstract Image dehazing has always been a challenging topic in image processing. The development of deep learning methods, especially the Generative Adversarial Networks(GAN), provides a new way for image dehazing. In recent years, many deep learning methods based on GAN have been applied to image dehazing. However, GAN has two problems in image dehazing. Firstly, For haze image, haze not only reduces the quality of the image, but also blurs the details of the image. For Gan network, it is difficult for the generator to restore the details of the whole image while removing the haze. Secondly, GAN model is defined as a minimax problem, which weakens the loss function. It is difficult to distinguish whether GAN is making progress in the training process. Therefore, we propose a Guided Generative Adversarial Dehazing Network(GGADN). Different from other generation adversarial networks, GGADN adds a guided module on the generator. The guided module verifies the network of each layer of the generator. At the same time, the details of the map generated by each layer are strengthened. Network training is based on the pre-trained VGG feature model and L1-regularized gradient prior which is developed by new loss function parameters. From the dehazing results of synthetic images and real images, proposed method is better than the state-of-the-art dehazing methods.

scholarly journals Unsupervised feature learning-based encoder and adversarial networks

2021 ◽  
Vol 8 (1) ◽  
Author(s):  
Endang Suryawati ◽  
Hilman F. Pardede ◽  
Vicky Zilvan ◽  
Ade Ramdan ◽  
Dikdik Krisnandi ◽  
...  
Keyword(s):  
Deep Learning ◽  
Feature Learning ◽  
Object Classification ◽  
Learning By Doing ◽  
Training Data ◽  
Generative Adversarial Networks ◽  
Good Representation ◽  
Learning Methods ◽  
Adversarial Networks ◽  
Unsupervised Deep Learning

AbstractIn this paper, we propose a novel deep learning-based feature learning architecture for object classification. Conventionally, deep learning methods are trained with supervised learning for object classification. But, this would require large amount of training data. Currently there are increasing trends to employ unsupervised learning for deep learning. By doing so, dependency on the availability of large training data could be reduced. One implementation of unsupervised deep learning is for feature learning where the network is designed to “learn” features automatically from data to obtain good representation that then could be used for classification. Autoencoder and generative adversarial networks (GAN) are examples of unsupervised deep learning methods. For GAN however, the trajectories of feature learning may go to unpredicted directions due to random initialization, making it unsuitable for feature learning. To overcome this, a hybrid of encoder and deep convolutional generative adversarial network (DCGAN) architectures, a variant of GAN, are proposed. Encoder is put on top of the Generator networks of GAN to avoid random initialisation. We called our method as EGAN. The output of EGAN is used as features for two deep convolutional neural networks (DCNNs): AlexNet and DenseNet. We evaluate the proposed methods on three types of dataset and the results indicate that better performances are achieved by our proposed method compared to using autoencoder and GAN.

Design Automation by Integrating Generative Adversarial Networks and Topology Optimization

2018 ◽  
Author(s):  
Sangeun Oh ◽  
Yongsu Jung ◽  
Ikjin Lee ◽  
Namwoo Kang
Keyword(s):  
Deep Learning ◽  
Topology Optimization ◽  
Unsupervised Learning ◽  
Design Automation ◽  
Generative Adversarial Networks ◽  
Adversarial Networks ◽  
And Topology ◽  
Recent Advances ◽  
Output Quality ◽  
Human Interventions

Recent advances in deep learning enable machines to learn existing designs by themselves and to create new designs. Generative adversarial networks (GANs) are widely used to generate new images and data by unsupervised learning. Certain limitations exist in applying GANs directly to product designs. It requires a large amount of data, produces uneven output quality, and does not guarantee engineering performance. To solve these problems, this paper proposes a design automation process by combining GANs and topology optimization. The suggested process has been applied to the wheel design of automobiles and has shown that an aesthetically superior and technically meaningful design can be automatically generated without human interventions.

Seismic compressive sensing by generative inpainting network: Toward an optimized acquisition survey

2019 ◽  
Vol 38 (12) ◽  
pp. 923-933 ◽  
Author(s):  
Xiaoyang Rebecca Li ◽  
Nikolaos Mitsakos ◽  
Ping Lu ◽  
Yuan Xiao ◽  
Xing Zhao
Keyword(s):  
Deep Learning ◽  
Compressive Sensing ◽  
Computational Cost ◽  
Seismic Survey ◽  
Generative Adversarial Networks ◽  
Promising Alternative ◽  
Seismic Surveys ◽  
Adversarial Networks ◽  
Seismic Acquisition ◽  
Seismic Images

The use of deep learning models as priors for compressive sensing tasks presents new potential for inexpensive seismic data acquisition. Conventional recovery usually suffers from undesired artifacts, such as oversmoothing, and high computational cost. Generative adversarial networks (GANs) offer promising alternative approaches that can improve quality and reveal finer details. An appropriately designed Wasserstein GAN trained on several historical surveys and capable of learning the statistical properties of the seismic wavelet's architecture is proposed. The efficiency and precision of this model at compressive sensing are validated in three steps. First, the existence of a sparse representation with different compression rates for seismic surveys is studied. Then, nonuniform samplings are studied using the proposed methodology. Finally, a recommendation is proposed for a nonuniform seismic survey grid based on the evaluation of reconstructed seismic images and metrics. The primary goal of the proposed deep learning model is to provide the foundations of an optimal design for seismic acquisition without a loss in imaging quality. Along these lines, a compressive sensing design of a nonuniform grid over an asset in the Gulf of Mexico, versus a traditional seismic survey grid that collects data uniformly every few feet, is suggested, leveraging the proposed method.

scholarly journals Audio-Based Drone Detection and Identification Using Deep Learning Techniques with Dataset Enhancement through Generative Adversarial Networks

Sensors ◽  
2021 ◽  
Vol 21 (15) ◽  
pp. 4953
Author(s):  
Sara Al-Emadi ◽  
Abdulla Al-Ali ◽  
Abdulaziz Al-Ali
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Recurrent Neural Network ◽  
Learning Algorithms ◽  
Generative Adversarial Networks ◽  
Generative Adversarial Network ◽  
Adversarial Networks ◽  
Detection And Identification ◽  
Learning Techniques ◽  
The Impact

Drones are becoming increasingly popular not only for recreational purposes but in day-to-day applications in engineering, medicine, logistics, security and others. In addition to their useful applications, an alarming concern in regard to the physical infrastructure security, safety and privacy has arisen due to the potential of their use in malicious activities. To address this problem, we propose a novel solution that automates the drone detection and identification processes using a drone’s acoustic features with different deep learning algorithms. However, the lack of acoustic drone datasets hinders the ability to implement an effective solution. In this paper, we aim to fill this gap by introducing a hybrid drone acoustic dataset composed of recorded drone audio clips and artificially generated drone audio samples using a state-of-the-art deep learning technique known as the Generative Adversarial Network. Furthermore, we examine the effectiveness of using drone audio with different deep learning algorithms, namely, the Convolutional Neural Network, the Recurrent Neural Network and the Convolutional Recurrent Neural Network in drone detection and identification. Moreover, we investigate the impact of our proposed hybrid dataset in drone detection. Our findings prove the advantage of using deep learning techniques for drone detection and identification while confirming our hypothesis on the benefits of using the Generative Adversarial Networks to generate real-like drone audio clips with an aim of enhancing the detection of new and unfamiliar drones.

scholarly journals Independent attenuation correction of whole body [18F]FDG-PET using a deep learning approach with Generative Adversarial Networks

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Karim Armanious ◽  
Tobias Hepp ◽  
Thomas Küstner ◽  
Helmut Dittmann ◽  
Konstantin Nikolaou ◽  
...  
Keyword(s):  
Deep Learning ◽  
Attenuation Correction ◽  
Fdg Pet ◽  
Whole Body ◽  
Generative Adversarial Networks ◽  
Learning Approach ◽  
Adversarial Networks ◽  
18F Fdg

Combining generative adversarial networks and agricultural transfer learning for weeds identification

2021 ◽  
Vol 204 ◽  
pp. 79-89
Author(s):  
Borja Espejo-Garcia ◽  
Nikos Mylonas ◽  
Loukas Athanasakos ◽  
Eleanna Vali ◽  
Spyros Fountas
Keyword(s):  
Transfer Learning ◽  
Generative Adversarial Networks ◽  
Adversarial Networks

Deep Learning Based Classification of Time Series of Chaotic Systems over Graphic Images

2021 ◽  
Author(s):  
Süleyman UZUN ◽  
Sezgin KAÇAR ◽  
Burak ARICIOĞLU
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Transfer Learning ◽  
Chaotic Systems ◽  
Initial Conditions ◽  
Step Size ◽  
Learning Methods ◽  
Data Set ◽  
Graphic Images ◽  
Different Chaotic Systems

Abstract In this study, for the first time in the literature, identification of different chaotic systems by classifying graphic images of their time series with deep learning methods is aimed. For this purpose, a data set is generated that consists of the graphic images of time series of the most known three chaotic systems: Lorenz, Chen, and Rossler systems. The time series are obtained for different parameter values, initial conditions, step size and time lengths. After generating the data set, a high-accuracy classification is performed by using transfer learning method. In the study, the most accepted deep learning models of the transfer learning methods are employed. These models are SqueezeNet, VGG-19, AlexNet, ResNet50, ResNet101, DenseNet201, ShuffleNet and GoogLeNet. As a result of the study, classification accuracy is found between 96% and 97% depending on the problem. Thus, this study makes association of real time random signals with a mathematical system possible.

Sign in / Sign up

Export Citation Format

Share Document