Deep Convolutional Generative Adversarial Network and Convolutional Neural Network for Smoke Detection

Generative models have been applied in many fields and can be evaluated with many methods. In the evaluation of generative models, the proper evaluation metric varies with the application field. Therefore, the evaluation of generative adversarial networks is inherently challenging. In this study, conditional deep convolutional generative adversarial networks were applied in mechanical fault diagnosis and then evaluated. We proposed three evaluation metrics of conditional deep convolutional generative adversarial networks: Jensen-Shannon divergence, kernel maximum mean discrepancy, and the 1-nearest neighbor classifier which were used to distinguish generated samples from real samples, test mode collapsing and detect overfitting based on the dataset of Electronic Engineering Laboratory of Case Western Reserve University and the planetary gearbox dataset measured in the laboratory. The Jensen-Shannon divergence could not well distinguish generated samples from real samples. However, the two metrics (maximum mean discrepancy and 1-nearest neighbor classifier) well-distinguished generated samples from real samples, thus verifying the applicability of conditional deep convolutional generative adversarial networks in the field of mechanical diagnosis.

Training and testing unmanned vehicles need various real data. However, data of some special or dangerous testing environment may not be accessible, or may only be accessible at certain times. So, using generative adversarial networks to learn the real traffic scene and generate a new scene is an effective way of solving the problem. In this paper, a framework of deep convolutional generative adversarial networks (DCGAN) was used to generate new traffic scene images and videos. Firstly, 300 sets of videos and images of overtaking scenes were selected as training data. A generator with convolutional neural network was used to generate samples. Then the training data and the generated samples were trained in a two-class discriminator that provides the probabilities of the samples that come from the generated sample and the training data respectively. Then the generator was updated by back propagation algorithm. Afterward, a new sample was generated and trained in the discriminator again. Repeated several times, a serial of generated samples were generated, the probability distribution of which is basically the same as the training data. The experiments show that this method can effectively generate realistic traffic scene images and videos.

Aiming at the issues of low individual identification accuracy of radiation sources under the condition of little samples, this paper proposes a way of individual identification of radiation source based on strengthening global deep convolutional generative adversarial network (SGDCGAN) and deep convolutional neural network (DCNN) to achieve data augmentation. The method first performs IQ map feature splicing processing on the input signal, and then uses DCNN to automatically obtain the deep essential features of the data. Besides, an adaptive improvement is created to the deep convolutional generative adversarial network, and a self‐attention mechanism is introduced into the discriminator and the generator to reinforce the integrity and authenticity of the generated samples. For the common gradient disappearance during model training, the gradient penalty mechanism and spectral normalization are added to make the training process more stable. Through comparison experiments on the collected ADS‐B signals, the experimental results have demonstrated that when the signal‐to‐noise ratio is 0 dB and the number of original samples in each class is 40, the recognition accuracy is improved by 23.6% after doubling the data. Compared with the DCGAN‐DCNN and GAN‐DCNN methods, the recognition accuracy is improved by 3.5% and 4%.

The fault diagnosis of hydrogen sensors is of great significance. However, it is difficult to collect data samples for some modes of hydrogen sensor signals, so the data samples may be unbalanced, which can seriously affect the fault diagnosis results. In this paper, we present a novel convolutional neural network (CNN)-based deep convolutional generative adversarial network (DCG) method (DCG-CNN) for gas sensor fault diagnosis. First, we transform the 1D fault signals of the gas sensor into 2D gray images for end-to-end conversion with no signal data information loss. Second, we use the DCG to enrich the 2D gray images of small fault data samples, which results in balanced sensor fault datasets. Third, we use the CNN method to improve the accuracy of fault diagnosis. In order to understand the internal mechanism of the CNN, we further visualize the learned feature maps of fault data samples in each layer of the CNN and try to analyze the reasons for the method's high performance. The fault diagnosis accuracy of the DCG-CNN is shown to be higher than that of other traditional methods.

Convolutional Neural Network (CNN) tend to have better results on large data sets and poor performance on small data sets, so the data augmentation is crucial for a CNN to get better performance based on the dataset with limited size. In this paper, Deep Convolution Generative Adversarial Network (DCGAN) was used to augment data to make the AlexNet perform better on an image classification task with small data sets. AlexNet was trained on a small anime face training set with only 160 samples to determine whether the anime face was male or female, and then tested its accuracy on a test set with 240 sample. Then, a pre-trained DCGAN was transferred to train on the male and female training sets respectively. And 2 DCGANs were obtained, one could generate male cartoon faces and another could generate female cartoon faces. The images generated by DCGANs were put in train set, which was used to train AlexNet again and the result was recorded. Other data augmentation methods such as cutout, cutmix and Noise Injection were compared as well. Finally, it is found that AlexNet has the best performance when using the DCGAN augmentation method, which can significantly improve the verification accuracy of the model.

Deep convolutional generative adversarial networks for the generation of numerous artificial spectrum‐compatible earthquake accelerograms using a limited number of ground motion records

In this paper, we propose a novel method for recognizing occluded offline handwritten Chinese characters based on deep convolutional generative adversarial network (DCGAN) and improved GoogLeNet. Different from previous methods, our proposed method is capable of inpainting and recognizing occluded characters without needing to know the concrete positions of corrupted regions. First, the generator and discriminator of DCGAN are combined to generate realistic Chinese characters from corrupted images, and the contextual loss and the content loss are further used to inpaint generated images. Finally, we use the improved GoogLeNet with traditional feature extraction methods to recognize the recovered handwritten Chinese characters. The proposed method is evaluated on the extended CASIA-HWDB1.1 dataset for two challenging inpainting tasks with different portions of blocks or random missing pixels. Experimental results show that our method can achieve higher repair rates and higher recognition accuracies than most of existing methods.

As a key safety component of automobiles, automobile steering knuckles must be subjected to strict quality control. Currently, the identification of cracks in finished products primarily relies on manual identification of fluorescent penetrant detection. Owing to the complex shape of the workpiece, the interference of the displayed image and the small sample size, the accuracy of the automatic discrimination result of the fluorescent penetrant detection image is directly reduced. Therefore, this study proposed a data augmentation method based on deep convolutional generative adversarial networks (DCGAN) for crack identification in automotive steering knuckle fluorescent penetration inspection images. An image acquisition platform was built for fluorescence penetration detection of automobile steering knuckles, and fluorescence display images of various parts of the workpiece were collected. Based on the feature analysis of the displayed image, the image was preprocessed to suppress relevant interference and extract crack candidate regions. Further, using the original crack image to train DCGAN, several crack image samples were generated, the ResNet network was trained with the expanded dataset, and the extracted candidate regions were identified. Finally, the experimental results show that the recall rate of the crack recognition method used in this paper is 95.1%, and the accuracy rate is 90.8%, which can better identify the crack defects in the fluorescent penetrant inspection image, compared with the non-generative data enhancement method.

Prediction of multi-layer metasurface design using conditional deep convolutional generative adversarial networks

Due to the real working conditions, the collected mechanical fault datasets are actually limited and always highly imbalanced, which restricts the diagnosis accuracy and stability. To solve these problems, we present an imbalanced fault diagnosis method based on the generative model of conditional-deep convolutional generative adversarial network (C-DCGAN) and provide a study in detail. Deep convolutional generative adversarial network (DCGAN), based on traditional generative adversarial networks (GAN), introduces convolutional neural network into the training for unsupervised learning to improve the effect of generative networks. Conditional generative adversarial network (CGAN) is a conditional model obtained through introducing conditional extension into GAN. C-DCGAN is a combination of DCGAN and CGAN. In C-DCGAN, based on the feature extraction ability of convolutional networks, through the structural optimization, conditional auxiliary generative samples are used as augmented data and applied in machine fault diagnosis. Two datasets (Bearing dataset and Planetary gear box dataset) are carried out to validate. The simulation experiments showed that the improved performance is mainly due to the generated signals from C-DCGAN to balance the dataset. The proposed method can deal with imbalanced fault classification problem much more effectively. This model could improve the accuracy of fault diagnosis and the generalization ability of the classifier in the case of small samples and display better fault diagnosis performance.

In order to improve the image quality of a specific class of crack images, as well as to solve the problems of insufficient size of the number of crack datasets and small number of complex crack images, a crack image generation model based on DCGAN (Deep Convolutional Generative Adversarial Network, DCGAN) is proposed, which has superior training stability and convergence speed. The experimental results show that DCGAN can generate a large number of real crack images with complex backgrounds more reliably than traditional image augmentation methods, effectively solving the problem of lack of crack images in special cases and greatly reducing the cost of crack image acquisition tasks.

Fire detection technologies remain a critical component of building automation. With the recent significant advances in computer vision, visual fire detection methods have been developed and integrated into building surveillance systems. Overfitting and accuracy challenges remain in fire detection when training datasets are limited. In this work, we tackle these challenges by developing a deep convolutional generative adversarial network (DCGAN) for highly accurate visual fire detection when training images are limited. Our model addresses three types of errors in visual fire detection with small training datasets: model overfitting, fire probability overestimation, and fire probability underestimation. The DCGAN includes a generator of fake fire images for self-supervised learning (SSL) and a discriminator for image classification. We designed computational experiments with high-quality datasets to test and validate our model against other supervised learning approaches. We also benchmarked the performance of the DCGAN against a best-in-class deep visual fire detection model. The results show that our model significantly outperforms other fire detection models on all performance metrics when trained with the same small dataset. The results demonstrate that the DCGAN effectively mitigates the three types of error when the training dataset is limited.

Reconstruction of the meso-scale concrete model using a deep convolutional generative adversarial network (DCGAN)

The COVID-19 pandemic has exerted a widespread influence on a global scale, leading numerous nations to prepare for the endemicity of COVID-19. The polymerase chain reaction (PCR) swab test has emerged as the prevailing technique for identifying viral infections within the current pandemic. Following this, the application of chest X-ray imaging in individuals provides an alternate approach for evaluating the existence of viral infection. However, it is imperative to further boost the quality of collected chest pictures via additional data augmentation. The aim of this paper is to provide a technique for the automated analysis of X-ray pictures using server processing with a deep convolutional generative adversarial network (DCGAN). The proposed methodology aims to improve the overall image quality of X-ray scans. The integration of deep learning with Xtreme Gradient Boosting in the DCGAN technique aims to improve the quality of X-ray pictures processed on the server. The training model employed in this work is based on the Inception V3 learning model, which is combined with XGradient Boost. The results obtained from the training procedure were quite interesting: the training model had an accuracy rate of 98.86%, a sensitivity score of 99.1%, and a recall rate of 98.7%.

Gastric cancer has a high mortality rate worldwide, but it can be prevented with early detection through regular gastroscopy. Herein, we propose a deep learning-based computer-aided diagnosis (CADx) system applying data augmentation to help doctors classify gastroscopy images as normal or abnormal. To improve the performance of deep learning, a large amount of training data are required. However, the collection of medical data, owing to their nature, is highly expensive and time consuming. Therefore, data were generated through deep convolutional generative adversarial networks (DCGAN), and 25 augmentation policies optimized for the CIFAR-10 dataset were implemented through AutoAugment to augment the data. Accordingly, a gastroscopy image was augmented, only high-quality images were selected through an image quality-measurement method, and gastroscopy images were classified as normal or abnormal through the Xception network. We compared the performances of the original training dataset, which did not improve, the dataset generated through the DCGAN, the dataset augmented through the augmentation policies of CIFAR-10, and the dataset combining the two methods. The dataset combining the two methods delivered the best performance in terms of accuracy (0.851) and achieved an improvement of 0.06 over the original training dataset. We confirmed that augmenting data through the DCGAN and CIFAR-10 augmentation policies is most suitable for the classification model for normal and abnormal gastric endoscopy images. The proposed method not only solves the medical-data problem but also improves the accuracy of gastric disease diagnosis.