检索结果-内蒙古大学图书馆

Suitable and Style-Consistent Multi-Texture Recommendation for Cartoon Illustrations

ACM Transactions on Multimedia Computing, Communications and Applications 2024年第7期20卷 1-26页

作者： Wu, Huisi Wang, Zhaoze Li, Yifan Liu, Xueting Lee, Tong-Yee College of Computer Science and Software Engineering Shenzhen University No. 3688 Nanhai Road Nanshan District Guangdong Shenzhen518060 China Dept. of Computer Science and Information Engineering National Cheng-Kung University No. 1 University Road Tainan70101 Taiwan

Texture plays an important role in cartoon illustrations to display object materials and enrich visual experiences. Unfortunately, manually designing and drawing an appropriate texture is not easy even for proficient artists, let alone novice or amateur people. While there exist tons of textures on the Internet, it is not easy to pick an appropriate one using traditional text-based search engines. Although several texture pickers have been proposed, they still require the users to browse the textures by themselves, which is still labor-intensive and time-consuming. In this article, an automatic texture recommendation system is proposed for recommending multiple textures to replace a set of user-specified regions in a cartoon illustration with visually pleasant look. Two measurements, the suitability measurement and the style-consistency measurement, are proposed to make sure that the recommended textures are suitable for cartoon illustration and at the same time mutually consistent in style. The suitability is measured based on the synthesizability, cartoonity, and region fitness of textures. The style-consistency is predicted using a learning-based solution since it is subjective to judge whether two textures are consistent in style. An optimization problem is formulated and solved via the genetic algorithm. Our method is validated on various cartoon illustrations, and convincing results are obtained. © 2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Textures

来源：评论

学校读者我要写书评

暂无评论

Enhanced Classification of Alzheimer's Disease Using Deep Learning and Machine Learning Models with MRI Scans

Enhanced Classification of Alzheimer's Disease Using Deep Le...

引用

2024 International Conference on Frontiers of Information Technology, FIT 2024

作者： Salahudin, Amna Qadeer, Eeman Ahmed, Ayesha Fahad, Labiba Gillani Dept. of Software Engineering FAST National University of Computer and Emerging Sciences Islamabad Pakistan Dept. of Computer Science FAST National University of Computer and Emerging Sciences Islamabad Pakistan

ISBN: (纸本)9798331510503

Alzheimer's disease (AD) is a slowly progressive neurodegenerative disease, which necessitates early and accurate diagnosis for effective management. This research presents a comprehensive approach to the classification of Alzheimer's disease using Magnetic Resonance Imaging (MRI) scans, focusing on four distinct classes: nondemented, mild, very mild, and moderate demented. Our main goals are to combine diverse datasets to improve the robustness and adaptability of the model and to fully investigate Alzheimer's disease in all four stages. To achieve these goals, we employed a combination of machine learning and deep learning models. The models utilized include Histogram of Oriented Gradients (HOG) with Support Vector Machine (SVM), Gabor filters with SVM, Gabor-Gray Level Co-occurrence Matrix (GLCM) with SVM, ResNet50, AlexNet, and InceptionV3. Among these, the InceptionV3 model achieved the highest accuracy of 96.20% using 5-fold cross-validation followed the Resnet50 model achieving accuracy of 95.43%. Our approach addresses key challenges such as class imbalance, high intra-class variation, inter-class similarities, and dataset variability, demonstrating significant potential in improving Alzheimer's disease classification and detection. © 2024 IEEE.

关键词： Gabor filters

来源：评论

学校读者我要写书评

暂无评论

GPU Accelerated Modelling and Real-time Rendering of Fluid Motion 38

GPU Accelerated Modelling and Real-time Rendering of Fluid M...

引用

38th International Conference on Image and Vision Computing New Zealand, IVCNZ 2023

作者： Valentine, William Mukundan, Ramakrishnan University of Canterbury Dept. Computer Science and Software Engineering Christchurch New Zealand

ISBN: (纸本)9798350370515

This paper proposes a fluid rendering pipeline that uses OpenGL-4 shaders to employ the parallel processing capabilities of the GPU. The fluid's surface mesh is produced using tessellation shader stages where the input patches are assigned tessellation levels based on the fluid heightmap's curvature. The curvature is stored using a texture buffer object which allows access by shaders, thus allowing the tessellation calculations to be carried out in parallel. Use of this adaptive tessellation method increases both the simulation's framerate as well as its capacity to handle a greater number of primitives. Furthermore, it more optimally distributes the mesh's vertices to effectively increase the level of detail without using more primitives. Polygon culling using the geometry shader further optimises the number of primitives used to define the fluid surface. The Phong-Blinn model is used for surface lighting. We propose two GPU-based fluid surface flow visualisation methods. Texture buffer objects can be used to store and update a surface texture. Alternatively, particle positions are updated each frame using the geometry shader and stored in a buffer object using transform feedback. These flow visualisation techniques are particularly effective for communicating the swirling motion of vortices. © 2023 IEEE.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

Unravelling the Enigma of Polycystic Ovary Syndrome (PCOS): Using MLAlgorithms 2

Unravelling the Enigma of Polycystic Ovary Syndrome (PCOS): ...

引用

2nd International Conference on Disruptive Technologies, ICDT 2024

作者： Juneja, Sonam Bhati, Bhoopesh Singh Atwal, Shikha Anand, Monika Maiti, Souvik Chandigarh University Dept. of Computer Science and Engineering Gharuan India Iiit Dept. of Computer Science and Engineering Sonepat India

ISBN: (纸本)9798350371055

The difficulties and ramifications of PCOS, which affects a sizable portion of women who are of reproductive age, are discussed in this review. The article covers the various clinical manifestations of PCOS, how it affects both reproductive and non-reproductive health, and how it's linked to psychological distress and metabolic disorders. It highlights how critical lifestyle modifications, early detection, and accurate diagnosis are. Additionally, the study presents machine learning techniques for PCOS detection, demonstrating the effectiveness of techniques like Random Forests, CNN, and SVM. An innovative CDSS that uses the Red Deer Algorithm shows encouraging accuracy. The necessity of continued research, diversified datasets, and cooperative efforts to enhance PCOS detection at the nexus of technology and healthcare is highlighted in the abstract's conclusion. © 2024 IEEE.

关键词： Deep Learning KNN Machine Learning PCOS Random Forest SVM

来源：评论

学校读者我要写书评

暂无评论

Architectural Model of Deep Convolution Neural Network Approach for Brain Tumor Prediction

Architectural Model of Deep Convolution Neural Network Appro...

引用

2024 International Conference on Intelligent Computing and Sustainable Innovations in Technology, IC-SIT 2024

作者： Panigrahi, Raghunath Sahu, Rekha Nayak, Suvendu Chandan Kabat, Manas Ranjan IMIT Dept. of Computer Science and Engineering Cuttack India Silicon University Dept. of Computer Science and Engineering Bhubaneswar India

ISBN: (纸本)9798350369175

Magnetic Resonance Imaging (MRI) is most commonly used technique to detect a tumor by physicians in medical science. The MRI images are three-dimensional images with the number of slices that are analyzed to identify the defective cells in the naked eye that leads to a time-consuming process. Machine Learning (ML) techniques are investigated as a solution to improve tumor identification in a better way. Early and accurate detection is crucial for effective treatment planning and improving patient outcomes. Traditionally, brain tumor diagnosis relies on manual analysis of MRI (Magnetic Resonance Imaging) scans by radiologists, which can be time-consuming and prone to human error. The advent of deep learning, particularly Convolutional Neural Networks (CNNs), has opened new avenues for automating and improving the accuracy of brain tumor prediction. In this paper, we study the Convolutional Neural Network(CNN) model which is popular in image analysis. Similar way, we develop a CNN model to process and classify the MRI images to detect the type of brain tumor. We examine the model's performance by using 5712 MRI images from a standard dataset available in the public domain. The aim of the work is to classify the tumor in four classes like: Glioma, Meningioma, Pituitary and No Tumour using the proposed CNN model. Our model follows data preprocessing, image encoding, filtering and classification. The proposed model scores 78% accuracy as its performance in the dataset which needs more investigations and comparison with other techniques in further study. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Secured Entry Control: An IoT and Deep Learning based Mask Detection System for COVID-19 Suspected Patient Screening and Access Regulation 3

Secured Entry Control: An IoT and Deep Learning based Mask D...

引用

3rd International Conference on Advancement in Electrical and Electronic engineering, ICAEEE 2024

作者： Sultana, Abida Peyal, Hasibul Islam Islam, Md. Injamul Abdal, Md. Nazmul Miraz, Shafiun Kontho, Md. Mahin Rahaman Leion, Zannatul Mawa Mamun, Abdul Al Rajshahi University of Engineering & Technology Dept. of Electrical & Computer Engineering Rajshahi Bangladesh Uttara University Dept. of Computer Science & Engineering Dhaka Bangladesh University of Liberal Arts Bangladesh Dept. of Computer Science and Engineering Dhaka Bangladesh Rajshahi University of Engineering & Technology Dept. of Computer Science & Engineering Rajshahi Bangladesh Tejgaon College Dept. of Computer Science and Engineering Dhaka Bangladesh

ISBN: (纸本)9798350388282

The new coronavirus SARS-CoV-2, which triggered the COVID-19 pandemic, has had an unparalleled effect on economies, cultures, and world health. In response to the critical need for strict COVID-19 screening systems in public areas, this study presents a creative Secured Entry Control system. Detecting and controlling possible COVID-19 carriers attempting to enter the country is made possible by this system, which makes use of deep learning algorithms and IoT technology. The mask detection algorithm, MobileNetV2 model, has an exceptional validation accuracy of 98.96\%. The model's reliability is supported by performance evaluations using ROC curves, confusion matrix analysis, and an AUC value of 98.96\%, which is close to the optimal AUC score of 100\%. Because MobileNetV2 is well-suited for low-processing devices, it is easy to deploy it on Raspberry Pi, which helps to create an affordable system. Furthermore, by spotting increased body temperatures, a contactless temperature sensor improves the system's ability to identify possible COVID-19 carriers. The functioning of the system is confirmed by the working prototype that is presented in the Experimental Results section. The main goal of this research is to create an autonomous system that is affordable and selectively allows access to people who are less likely to transmit COVID-19 in public areas. To achieve the overall objective of reducing the spread of COVID-19 in public spaces, this research highlights the successful integration of mask identification algorithms and IoT. © 2024 IEEE.

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

Textual Authenticity in the AI Era: Evaluating BERT and RoBERTa with Logistic Regression and Neural Networks for Text Classification 16

Textual Authenticity in the AI Era: Evaluating BERT and RoBE...

引用

16th International Symposium on Electronics and Telecommunications, ISETC 2024

作者： Hazim, Layth Rafea Ata, Oguz Altinbas University Dept. of Electrical and Computer Engineering Istanbul Turkey College of Computer Science and Mathematics Tikrit University Tikrit Iraq Altinbas University Dept. of Software Engineering Istanbul Turkey

ISBN: (纸本)9798350390865

AI-generated content impersonating human writing is an issue that has gained attention as AI spreads its wings. This particular study serves as a comparison between the existing Logistic Regression and Feedforward Neural Networks (FNNs) by employing sentence-BERT-appended models and RoBERTa in the determination of authenticity in a given text. A dataset that was balanced between human-written and AI-generated texts was utilized. Techniques like tokenization and normalization were first used, followed by feature extraction using transformer-based models. Cross-validation and confusion matrix analysis that used measures such as accuracy, precision, recall, F1 score, and ROC AUC were included to guarantee the models' robustness. The hybrid RoBERTa-FNN model that was deposed challengers was the most outstanding model in respect of precision and recall, and the highest accuracy (99.95%) was obtained as mentioned in the data. The improved performance serves as a proof of how effectively RoBERTa uses its embeddings to represent context on the fine-grained level required for this kind of text classification. This work is a stepping stone to the creation of strong AI text detection systems, besides our advancements to the knowledge of models and embedding performance with respect to text classification. The results lay emphasis on the selection of model configuration and the embedding technique, as they are the key factors in achieving the best results in practical applications. © 2024 IEEE.

关键词： Logistic regression

来源：评论

学校读者我要写书评

暂无评论

An Effective Deep Learning Approach for Kidney Stone Detection in CT Scan Images using Convolutional Neural Network

An Effective Deep Learning Approach for Kidney Stone Detecti...

引用

2024 IEEE International Conference on Computing, Applications and Systems, COMPAS 2024

作者： Islam, Md. Nurul Khan Chowdhury, Md. Fuad Hasan Nusrat, Fariha University of Scholars Dept. of Computer Science and Engineering Dhaka Bangladesh United International University Dept. of Computer Science and Engineering Dhaka Bangladesh Jahangirnagar University Dept. of Computer Science and Engineering Dhaka Bangladesh

ISBN: (纸本)9798331529765

Kidney stones represent a significant global health concern, frequently causing individuals to seek immediate medical care due to intense pain. Radiological imaging modalities are one of the most common modes of diagnosis, but specialist interpretation is always important for accuracy. In this paper, we present computer vision in an automated kidney stone detection system that involves five stages: loading the dataset, preprocessing data, loading the InceptionV3 model, predicting, and evaluating performance. Various methods of preprocessing and data augmentation were utilized to make the model less biased and more robust. This study uses the InceptionV3 architecture to classify kidney stones in CT scan images of KUB (Kidney, Ureter, and Bladder). We utilize a total dataset of 433 participants, either diagnosed with kidney stones or in a normal state. 278 cases are diagnosed with kidney stones, and 165 are normal cases. The proposed approach achieved an impressive 98.55% accuracy across performance metrics when evaluated on an independent test dataset. © 2024 IEEE.

关键词： computerized tomography

来源：评论

学校读者我要写书评

暂无评论

Using LLMs for Annotation and ML Methods for Comparative Analysis of Large-Scale Turkish Sentiment Datasets 9

Using LLMs for Annotation and ML Methods for Comparative Ana...

引用

9th International Conference on computer science and engineering, UBMK 2024

作者： Ezin, Ercan Kiziltepe, Rukiye Savran Karakus, Murat Harran University Dept. of Computer Engineering Şanliurfa Turkey Karadeniz Tech. University Dept. of Software Engineering Trabzon Turkey Ankara University Dept. of Software Engineering Ankara Turkey

ISBN: (纸本)9798350365887

With the increasing prevalence of e-commerce plat-forms, understanding customer sentiments expressed in product reviews is crucial for assessing platform and product performance. However, traditional sentiment analysis (SA) models often struggle with low-resource languages such as Turkish, particularly due to the language's agglutinative nature and morphological complexity. To address this, several public datasets are now available online. However, these datasets are often annotated using rule-based methods, resulting in poor accuracy for Machine Learning (ML) models. This study addresses these challenges by employing two large Turkish datasets, TRSAv1 and VSCR, and improving their sentiment annotations through advanced Large Language Models (LLMs) such as ChatGPT4o-mini. The reannotation significantly enhanced the performance of traditional ML models-Logistic Regression, Naive Bayes, Random Forest, and Support Vector Machine-especially in the neutral and negative sentiment classes, as evidenced by increased precision, recall, and F1-scores. The findings underscore the importance of thorough data preparation and demonstrate that utilizing LLMs for sentiment annotation in Turkish can lead to more reliable SA, providing a stronger foundation for developing advanced SA tailored to low-resource languages. This study is the first to employ recent LLM for relabelling Turkish customer reviews, highlighting the potential for improved accuracy in SA for low-resource languages. © 2024 IEEE.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

Privacy-Preserving Speech Analysis- A Review 6

Privacy-Preserving Speech Analysis- A Review

引用

6th International Conference on Computational Intelligence and Communication Technologies, CCICT 2024

作者： Kumar, Pankaj Shukla, Mayank Malik, Medha Dept. of Computer Science and Engineering ABES Engineering College Ghaziabad India Dept. of Computer Science and Engineering ABES Engineering College Ghaziabad India

ISBN: (纸本)9798350374629

Speech-recognition technology is an essential component of many smart IoT devices. To make these systems secure and privacy-preserving, it is important to consider outsourcing speech-recognition functionality to a third-party storage provider. By removing the speech-recognition process from the device itself, the amount of personal data stored on the device is minimized. Additionally, the provider can ensure the strictest security standards are met when the audio data is stored, including the use of encryption and access control mechanisms. Furthermore, the provider can also ensure that the speech-recognition process is done using the latest technologies to ensure accurate recognition of the audio data. To ensure the end-user's privacy is kept intact, the provider should allow only authorized users to access the audio data. Additionally, the provider should also enable data access logging and audit trails allowing for the tracking of any access to the data. By outsourcing speech-recognition services to a third-party, smart IoT devices can ensure the highest levels of privacy and security while still enabling users to get the most out of their device. The server processes the encrypted data to generate recognition results, such as user commands and voice printing, without compromising the privacy of the user's voice data. To enhance voice data privacy, federated learning can be employed in the design of these models. This approach enables distributed training of the model without the need to transmit the user's voice data to a central server. The adoption of this approach in the IoT industry is gaining momentum as it offers the potential to create smarter and more secure IoT devices that inspire customer trust. ©2024 IEEE.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：