Search results

With the development of intelligent technology, deep learning has made significant progress and has been widely used in various fields. Deep learning is data-driven, and its training process requires a large amount of data to improve model performance. However, labeled data is expensive and not readily available.

To address the above problem, researchers have integrated semi-supervised and deep learning, using a limited number of labeled data and many unlabeled data to train models. In this paper, Generative Adversarial Networks (GANs) are analyzed as an entry point. Firstly, we discuss the current research on GANs in image super-resolution applications, including supervised, unsupervised, and semi-supervised learning approaches. Secondly, based on semi-supervised learning, different optimization methods are introduced as an example of image classification. Eventually, experimental comparisons and analyses of existing semi-supervised optimization methods based on GANs will be performed.

Following the analysis of the selected studies, we summarize the problems that existed during the research process and propose future research directions.

This paper reviews and analyzes research on generative adversarial networks for image super-resolution and classification from various learning approaches. The comparative analysis of experimental results on current semi-supervised GAN optimizations is performed to provide a reference for further research.

This study presents an ontological approach to assess the architectural outputs of generative adversarial networks. This paper aims to assess the performance of the generative adversarial network in representing building knowledge.

The proposed ontological assessment consists of five steps. These are, respectively, creating an architectural data set, developing ontology for the architectural data set, training the You Only Look Once object detection with labels within the proposed ontology, training the StyleGAN algorithm with the images in the data set and finally, detecting the ontological labels and calculating the ontological relations of StyleGAN-generated pixel-based architectural images. The authors propose and calculate ontological identity and ontological inclusion metrics to assess the StyleGAN-generated ontological labels. This study uses 300 bay window images as an architectural data set for the ontological assessment experiments.

The ontological assessment provides semantic-based queries on StyleGAN-generated architectural images by checking the validity of the building knowledge representation. Moreover, this ontological validity reveals the building element label-specific failure and success rates simultaneously.

This study contributes to the assessment process of the generative adversarial networks through ontological validity checks rather than only conducting pixel-based similarity checks; semantic-based queries can introduce the GAN-generated, pixel-based building elements into the architecture, engineering and construction industry.

To improve the speed and accuracy of turbine blade film cooling design process, the most advanced deep learning models were introduced into this study to investigate the most suitable define for prediction work. This paper aims to create a generative surrogate model that can be applied on multi-objective optimization problems.

The latest backbone in the field of computer vision (Swin-Transformer, 2021) was introduced and improved as the surrogate function for prediction of the multi-physics field distribution (film cooling effectiveness, pressure, density and velocity). The basic samples were generated by Latin hypercube sampling method and the numerical method adopt for the calculation was validated experimentally at first. The training and testing samples were calculated at experimental conditions. At last, the surrogate model predicted results were verified by experiment in a linear cascade.

The results indicated that comparing with the Multi-Scale Pix2Pix Model, the Swin-Transformer U-Net model presented higher accuracy and computing speed on the prediction of contour results. The computation time for each step of the Swin-Transformer U-Net model is one-third of the original model, especially in the case of multi-physics field prediction. The correlation index reached more than 99.2% and the first-order error was lower than 0.3% for multi-physics field. The predictions of the data-driven surrogate model are consistent with the predictions of the computational fluid dynamics results, and both are very close to the experimental results. The application of the Swin-Transformer model on enlarging the different structure samples will reduce the cost of numerical calculations as well as experiments.

The number of U-Net layers and sample scales has a proper relationship according to equation (8). Too many layers of U-Net will lead to unnecessary nonlinear variation, whereas too few layers will lead to insufficient feature extraction. In the case of Swin-Transformer U-Net model, incorrect number of U-Net layer will reduce the prediction accuracy. The multi-scale Pix2Pix model owns higher accuracy in predicting a single physical field, but the calculation speed is too slow. The Swin-Transformer model is fast in prediction and training (nearly three times faster than multi Pix2Pix model), but the predicted contours have more noise. The neural network predicted results and numerical calculations are consistent with the experimental distribution.

This paper creates a generative surrogate model that can be applied on multi-objective optimization problems. The generative adversarial networks using new backbone is chosen to adjust the output from single contour to multi-physics fields, which will generate more results simultaneously than traditional surrogate models and reduce the time-cost. And it is more applicable to multi-objective spatial optimization algorithms. The Swin-Transformer surrogate model is three times faster to computation speed than the Multi Pix2Pix model. In the prediction results of multi-physics fields, the prediction results of the Swin-Transformer model are more accurate.

This paper aims to improve the manufacturability of additive manufacturing (AM) for topology-optimized (TO) structures. Enhancement of manufacturability focuses on modifying geometric constraints and classifying the building orientation (BO) of AM parts to reduce stresses and support structures (SSs). To this end, artificial intelligence (AI) networks are being developed to automate design for additive manufacturing (DfAM).

This study considers three geometric constraints for their correction by convolutional autoencoders (CAEs) and transfer learning (TL). Furthermore, BOs of AM parts are classified using generative adversarial (GAN) and classification networks to reduce the SS. To verify the results, finite element analysis (FEA) is performed to compare the stresses of modified components with the original ones. Moreover, one sample is produced by the laser-based powder bed fusion (LB-PBF) in the BO predicted by the AI to observe its SSs.

CAE and TL resulted in promoting the manufacturability of TO components. FEA demonstrated that enhancing manufacturability leads to a 50% reduction in stresses. Additionally, training GAN and pre-training the ResNet-18 resulted in 80%, 95% and 96% accuracy for training, validation and testing. The production of a sample with LB-PBF demonstrated that the predicted BO by ResNet-18 does not require SSs.

This paper provides an automatic platform for DfAM of TO parts. Consequently, complex TO parts can be designed most feasibly and manufactured by AM technologies with minimal material usage, residual stresses and distortions.

Predicting student performance is crucial in educational settings to identify and support students who may need additional help or resources. Understanding and predicting student performance is essential for educators to provide targeted support and guidance to students. By analyzing various factors like attendance, study habits, grades, and participation, teachers can gain insights into each student’s academic progress. This information helps them tailor their teaching methods to meet the individual needs of students, ensuring a more personalized and effective learning experience. By identifying patterns and trends in student performance, educators can intervene early to address any challenges and help students acrhieve their full potential. However, the complexity of human behavior and learning patterns makes it difficult to accurately forecast how a student will perform. Additionally, the availability and quality of data can vary, impacting the accuracy of predictions. Despite these obstacles, continuous improvement in data collection methods and the development of more robust predictive models can help address these challenges and enhance the accuracy and effectiveness of student performance predictions. However, the scalability of the existing models to different educational settings and student populations can be a hurdle. Ensuring that the models are adaptable and effective across diverse environments is crucial for their widespread use and impact. To implement a student’s performance-based learning recommendation scheme for predicting the student’s capabilities and suggesting better materials like papers, books, videos, and hyperlinks according to their needs. It enhances the performance of higher education.

Thus, a predictive approach for student achievement is presented using deep learning. At the beginning, the data is accumulated from the standard database. Next, the collected data undergoes a stage where features are carefully selected using the Modified Red Deer Algorithm (MRDA). After that, the selected features are given to the Deep Ensemble Networks (DEnsNet), in which techniques such as Gated Recurrent Unit (GRU), Deep Conditional Random Field (DCRF), and Residual Long Short-Term Memory (Res-LSTM) are utilized for predicting the student performance. In this case, the parameters within the DEnsNet network are finely tuned by the MRDA algorithm. Finally, the results from the DEnsNet network are obtained using a superior method that delivers the final prediction outcome. Following that, the Adaptive Generative Adversarial Network (AGAN) is introduced for recommender systems, with these parameters optimally selected using the MRDA algorithm. Lastly, the method for predicting student performance is evaluated numerically and compared to traditional methods to demonstrate the effectiveness of the proposed approach.

The accuracy of the developed model is 7.66%, 9.91%, 5.3%, and 3.53% more than HHO-DEnsNet, ROA-DEnsNet, GTO-DEnsNet, and AOA-DEnsNet for dataset-1, and 7.18%, 7.54%, 5.43% and 3% enhanced than HHO-DEnsNet, ROA-DEnsNet, GTO-DEnsNet, and AOA-DEnsNet for dataset-2.

The developed model recommends the appropriate learning materials within a short period to improve student’s learning ability.

This study aims to use deep neural networks (DNNs) to learn the conduction heat transfer physics and estimate temperature distribution images in a physical domain without using any physical model or mathematical governing equation.

Two novel DNNs capable of learning the conduction heat transfer physics were defined. The first DNN (U-Net autoencoder residual network [UARN]) was designed to extract local and global features simultaneously. In the second DNN, a conditional generative adversarial network (CGAN) was used to enhance the accuracy of UARN, which is referred to as CGUARN. Then, novel loss functions, introduced based on outlier errors, were used to train the DNNs.

A UARN neural network could learn the physics of heat transfer. Within a few epochs, it reached mean and outlier errors that other DNNs could never reach after many epochs. The composite outlier-mean error as a loss function showed excellent performance in training DNNs for physical images. A UARN could excellently capture local and global features of conduction heat transfer, whereas the composite error could accurately guide DNN to extract high-level information by estimating temperature distribution images.

This study offers a unique approach to estimating physical information, moving from traditional mathematical and physical models to machine learning approaches. Developing novel DNNs and loss functions has shown promising results, opening up new avenues in heat transfer physics and potentially other fields.

The purpose of this study is to comprehensively examine the ethical implications surrounding generative artificial intelligence (AI).

Leveraging a novel methodological approach, the study curates a corpus of 364 documents from Scopus spanning 2022 to 2024. Using the term frequency-inverse document frequency (TF-IDF) and structural topic modeling (STM), it quantitatively dissects the thematic essence of the ethical discourse in generative AI across diverse domains, including education, healthcare, businesses and scientific research.

The results reveal a diverse range of ethical concerns across various sectors impacted by generative AI. In academia, the primary focus is on issues of authenticity and intellectual property, highlighting the challenges of AI-generated content in maintaining academic integrity. In the healthcare sector, the emphasis shifts to the ethical implications of AI in medical decision-making and patient privacy, reflecting concerns about the reliability and security of AI-generated medical advice. The study also uncovers significant ethical discussions in educational and financial settings, demonstrating the broad impact of generative AI on societal and professional practices.

This study provides a foundation for crafting targeted ethical guidelines and regulations for generative AI, informed by a systematic analysis using STM. It highlights the need for dynamic governance and continual monitoring of AI’s evolving ethical landscape, offering a model for future research and policymaking in diverse fields.

The study introduces a unique methodological combination of TF-IDF and STM to analyze a large academic corpus, offering new insights into the ethical implications of generative AI across multiple domains.

The aim of this study is to offer valuable insights to businesses and facilitate better understanding on transformer-based models (TBMs), which are among the widely employed generative artificial intelligence (GAI) models, garnering substantial attention due to their ability to process and generate complex data.

Existing studies on TBMs tend to be limited in scope, either focusing on specific fields or being highly technical. To bridge this gap, this study conducts robust bibliometric analysis to explore the trends across journals, authors, affiliations, countries and research trajectories using science mapping techniques – co-citation, co-words and strategic diagram analysis.

Identified research gaps encompass the evolution of new closed and open-source TBMs; limited exploration across industries like education and disciplines like marketing; a lack of in-depth exploration on TBMs' adoption in the health sector; scarcity of research on TBMs' ethical considerations and potential TBMs' performance research in diverse applications, like image processing.

The study offers an updated TBMs landscape and proposes a theoretical framework for TBMs' adoption in organizations. Implications for managers and researchers along with suggested research questions to guide future investigations are provided.

The purpose of this paper is to highlight the key technical challenges that derive from the recently proposed European Artificial Intelligence Act and specifically, to investigate the applicability of the requirements that the AI Act mandates to high-risk AI systems from the perspective of AI security.

This paper presents the main points of the proposed AI Act, with emphasis on the compliance requirements of high-risk systems. It matches known AI security threats with the relevant technical requirements, it demonstrates the impact that these security threats can have to the AI Act technical requirements and evaluates the applicability of these requirements based on the effectiveness of the existing security protection measures. Finally, the paper highlights the necessity for an integrated framework for AI system evaluation.

The findings of the EU AI Act technical assessment highlight the gap between the proposed requirements and the available AI security countermeasures as well as the necessity for an AI security evaluation framework.

AI Act, high-risk AI systems, security threats, security countermeasures.

This paper aims to explore recent advances in drift compensation algorithms for Electronic Nose (E-nose) technology and addresses sensor drift challenges through offline, online and neural network-based strategies. It offers a comprehensive review and covers causes of drift, compensation methods and future directions. This synthesis provides insights for enhancing the reliability and effectiveness of E-nose systems in drift issues.

The article adopts a comprehensive approach and systematically explores the causes of sensor drift in E-nose systems and proposes various compensation strategies. It covers both offline and online compensation methods, as well as neural network-based approaches, and provides a holistic view of the available techniques.

The article provides a comprehensive overview of drift compensation algorithms for E-nose technology and consolidates recent research insights. It addresses challenges like sensor calibration and algorithm complexity, while discussing future directions. Readers gain an understanding of the current state-of-the-art and emerging trends in electronic olfaction.

This article presents a comprehensive review of the latest advancements in drift compensation algorithms for electronic nose technology and covers the causes of drift, offline drift compensation algorithms, online drift compensation algorithms and neural network drift compensation algorithms. The article also summarizes and discusses the current challenges and future directions of drift compensation algorithms in electronic nose systems.