Article
Geosciences, Multidisciplinary
Xingye Liu, Jiwei Cheng, Yue Cai, Qianwen Mo, Chao Li, Shaohuan Zu
Summary: Sedimentary facies simulation is an important task in sedimentary environment analysis and reservoir characterization. Traditional methods have limitations, so we developed an intelligent method for automatic simulation. The method utilizes deep convolutional generative adversarial networks and an improved image quilting algorithm to generate simulation results. Model testing confirms the effectiveness and reliability of the method, which has also been successfully applied to non-stationary geological facies simulation.
MARINE AND PETROLEUM GEOLOGY
(2022)
Article
Computer Science, Interdisciplinary Applications
Rachel Xu, Vladimir Puzyrev, Chris Elders, Ebrahim Fathi Salmi, Ewan Sellers
Summary: This article explores the potential of using Generative Adversarial Networks (GANs) to augment seismic facies classification through training Convolutional Neural Networks (CNNs). The study finds that increasing the size of the training dataset improves the accuracy of the CNN classification, and a balance between diversity and consistency is important for optimal performance.
COMPUTERS & GEOSCIENCES
(2023)
Article
Engineering, Electrical & Electronic
Mu-Yen Chen, Wei Wei, Han-Chieh Chao, Yi-Fen Li
Summary: Robotic musicianship research focuses on enabling robots to analyze, reason, and generate music autonomously, aiming to achieve inspiring and meaningful musical interactions between humans and artificially creative robots. This research introduces a new model of automatic music generation using least squares and generative adversarial networks (GANs). By applying sequence generation adversarial network (SeqGAN) techniques, the research addresses the challenges in generating classical piano melodies. The proposed method, called Least Squares SeqGAN (LS-SeqGAN), effectively creates melody units on different chords and generates a set of music pieces with high quality and creativity, offering a robust infrastructure for human-robotic interaction.
IEEE SENSORS JOURNAL
(2022)
Article
Computer Science, Information Systems
Aurele Tohokantche Aurele Tohokantche, Wenming Cao, Xudong Mao, Si Wu, Hau-San Wong, Qing Li
Summary: This paper proposes a parametric and robust AB loss function to improve the performance of generative adversarial networks (GAN) on different datasets and alleviate the issue of mode collapse. Experimental results demonstrate that this approach can enhance the quality of synthetic images.
INFORMATION SCIENCES
(2022)
Article
Geosciences, Multidisciplinary
Fleford Redoloza, Liangping Li, Arden Davis
Summary: Groundwater-flow and contaminant-transport modeling can benefit from the use of machine learning techniques such as generative adversarial networks (GANs). This study focuses on a progressive growing GAN (PGGAN) to generate geologically realistic images of channel aquifers based on field observations. The conditioning behavior and its influence on the network architecture were measured using a novel metric called the conditioning ratio. The results revealed different conditioning behaviors based on the number of conditioning arrays injected into the generator.
HYDROGEOLOGY JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Aurele Tohokantche Gnanha, Wenming Cao, Xudong Mao, Si Wu, Hau-San Wong, Qing Li
Summary: In this paper, we propose the Residual Generator for GAN (Rg-GAN) to bridge the gap between theory and practice in GAN, by minimizing the residual between the loss of the generated data and the loss of the generated data from the perspective of the discriminator. The experiments show that Rg-GAN is robust to mode collapse and improves the generation quality of GAN in terms of FID and IS scores.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Information Systems
Mahsa Soleimani, Ali Nazari, Mohsen Ebrahimi Moghaddam
Summary: DeepFake involves the use of deep learning and artificial intelligence techniques to produce or change video and image contents. It can lead to fake news, crimes, and affect facial recognition systems. This study presents a deep learning approach using the entire face and face patches to distinguish real/fake images, overcoming limitations like blurring and obstruction. Experimental results show that this approach performs better and weighing the patches improves accuracy.
MULTIMEDIA SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Gaoming Yang, Mingwei Li, Xianjing Fang, Ji Zhang, Xingzhu Liang
Summary: Adversarial examples pose a security threat to deep learning models, and the proposed Attack Without a Target Model (AWTM) method achieves high attack success rate with low time cost.
PEERJ COMPUTER SCIENCE
(2021)
Article
Computer Science, Information Systems
Tao Bai, Jinqi Luo, Jun Zhao
Summary: Deep-learning-based image-recognition systems on mobile devices are vulnerable to adversarial examples. To tackle this issue, this study proposes a method for generating inconspicuous adversarial patches with one single image. The patches are produced in a coarse-to-fine manner and encouraged to be consistent with the background images while retaining strong attack abilities. The approach demonstrates high attack success rates in white-box and black-box settings, with minimal risk of being detected and evading human observations.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Computer Science, Interdisciplinary Applications
Fei Hu, Chunlei Wu, Jiangwei Shang, Yiming Yan, Leiquan Wang, Huan Zhang
Summary: This paper explores a flexible approach to generate synthetic realizations with desired geological styles by introducing numerical codes and extending the existing mapping model. The resulting generative model can synthesize images that respect both hard data and exhibit specific geological styles.
COMPUTERS & GEOSCIENCES
(2023)
Article
Computer Science, Information Systems
Ivan Fontana, Marc Langheinrich, Martin Gjoreski
Summary: Human mobility modeling is essential for various societal aspects, such as disease transmission modeling and urban planning. The application of deep learning to human mobility has been facilitated by the availability of vast mobility data. This study explores cutting-edge methods for trajectory generation, classification, and next-location prediction, and proposes a privacy-aware approach for predicting next-week trajectories by combining a Generative Adversarial Network and a deep learning model for user identification. Experimental results show that the generated trajectories preserve privacy without significant deviation from the original ones.
Article
Computer Science, Hardware & Architecture
Zuobin Xiong, Zhipeng Cai, Chunqiang Hu, Daniel Takabi, Wei Li
Summary: Recent progress has demonstrated the success of neural networks in various emerging applications, but the research on utilizing neural networks for secure communication is lacking. The existing neural network-based communication system has critical security flaws. This article investigates the vulnerabilities of the current system and proposes attack models and a new defense mechanism to improve security performance. The effectiveness of the proposed neural communication system is shown through theoretical proof and comprehensive real data experiments.
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Wonkeun Jo, Dongil Kim
Summary: In this study, a new oversampling method called OBGAN is proposed to address the class imbalance issue by considering the relationship between minority and majority classes. OBGAN uses independent discriminators to competitively affect the training of the generator, allowing it to capture different regions of minority and majority classes while avoiding mode collapse problem.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Geography, Physical
Matthew Yates, Glen Hart, Robert Houghton, Mercedes Torres Torres, Michael Pound
Summary: Concerns have arisen about the authenticity of images due to advanced image generation techniques. This study focuses on the generation and evaluation of Earth Observation (EO) data using state-of-the-art GAN models. The synthesized EO images are found to deceive humans, and current performance metrics have limitations in quantifying visual quality.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2022)
Article
Computer Science, Artificial Intelligence
Yiqiang Wu, Dapeng Tao, Yong Luo, Jun Cheng, Xuelong Li
Summary: In this paper, a novel frame-level face anti-spoofing method, Covered Style Mining-GAN (CSM-GAN), is proposed, which converts face anti-spoofing detection into a style transfer process without any prior information. Comprehensive experiments show that the proposed method outperforms current state-of-the-art and produces better visual diversity in difference maps.
PATTERN RECOGNITION
(2022)
Article
Multidisciplinary Sciences
Yuyang Liu, Zhaoliang Li, Mao Pan
Article
Metallurgy & Metallurgical Engineering
Yu-yang Liu, Mao Pan, Shi-qi Liu
JOURNAL OF CENTRAL SOUTH UNIVERSITY
(2019)
Article
Energy & Fuels
Jianpeng Yao, Qingbin Liu, Wenling Liu, Yuyang Liu, Xiaodong Chen, Mao Pan
Article
Energy & Fuels
Shiqi Liu, Yuyang Liu, Xiaowei Zhang, Wei Guo, Lixia Kang, Rongze Yu, Yuping Sun
Summary: This paper introduces a method for evaluating shale gas sweet spots using fuzzy mathematics. By considering multiple parameters, the geological and engineering sweet spots in the target area can be evaluated quickly and effectively. The results show high rationality and accuracy, providing valuable assistance in well-pattern deployment and fracture design for shale gas production.
Article
Geochemistry & Geophysics
Shiqi Liu, Yuyang Liu
Summary: This study examines the sedimentary facies of the Jurassic formations in the Wuerhe area of the Junggar Basin, which is a key area for oil and gas exploration. By analyzing well logs and other data, the study reveals the sedimentary evolution of the area and concludes that the Jurassic system in the Wuerhe area is a fan delta-lacustrine-fan delta sedimentary system.
Review
Chemistry, Multidisciplinary
Yuyang Liu, Xiaowei Zhang, Wei Guo, Lixia Kang, Jinliang Gao, Rongze Yu, Yuping Sun, Mao Pan
Summary: This article comprehensively analyzes the grid generation and property interpolation methods in three-dimensional geological property modeling. It is found that in numerical simulation of oil reservoirs, the orthogonal hexahedral grid remains the most suitable grid model. For interpolation methods, most geological phenomena are nonstationary, and the main development trend is to increase geological constraints and reduce the limitation of stationarity.
APPLIED SCIENCES-BASEL
(2022)
Review
Geochemistry & Geophysics
Xun Gong, Xinhua Ma, Yuyang Liu, Guanfang Li
Summary: This paper systematically discusses the current status and future development direction of artificial fracture propagation law in shale reservoirs. The research methods include indoor physical simulation experiments and numerical simulation methods. The former has the advantages of simple operation and intuitive image, while the latter improves the accuracy of the model solution through integrating numerical algorithms. The propagation law of artificial fractures is influenced by geological factors and engineering factors, and research should be based on geological factors and optimize engineering influencing factors.