Facial recognition technology is widely used for biometric authentication, but it struggles with accurately identifying individuals from low-resolution images. This study introduces a new method for improving low-reso...
详细信息
ISBN:
(纸本)9798350366396;9798350366389
Facial recognition technology is widely used for biometric authentication, but it struggles with accurately identifying individuals from low-resolution images. This study introduces a new method for improving low-resolution face recognition (LRFR) by combining super-resolution techniques and multilinear subspace learning. Using deep learning VGG-face models for feature extraction and super-resolution methods like SRGAN and SRResNet, the proposed method, Multilinear Side-information-based Discriminant Analysis (MSIDA), enhances facial image quality. This methodology integrates image enhancement, deep feature extraction, and tensor subspace learning for precise face verification. The effectiveness of this approach is demonstrated on the Labeled Faces in the Wild (LFW) dataset, showing superior performance in face verification tasks compared to existing methods. MSIDA achieves accuracies of 91.03%, 91.60%, and 92.57% for different resolutions. This advancement represents significant progress in image processing and pattern recognition, particularly for scenarios involving low-resolution images.
Quantum computing is a technology that is frantically growing. It revolutionizes many industries for all different types of things and will solve problems that transpositional computers have not thought of yet. Quantu...
详细信息
Radio Frequency Identification (RFID) is a technology that is used widely for mobile payment, the security is the main issue of the usage of this technology. The aim of this study is to identify the key factors influe...
详细信息
The theory of discourse functional pragmatics posits that every discourse unit serves a purpose, and functional pragmatics can reveal its role in the text. To achieve a more informative representation of paragraphs, m...
详细信息
ISBN:
(纸本)9789819756711;9789819756728
The theory of discourse functional pragmatics posits that every discourse unit serves a purpose, and functional pragmatics can reveal its role in the text. To achieve a more informative representation of paragraphs, mitigate the long-tail problem, and incorporate the specificity of different models, we propose a Chinese discourse Functional Pragmatics Recognition model based on Multi-level information and Ensemble learning (FPRME). Specifically, two different encoding methods are combined to obtain paragraph representation that contains rich word-level information. Then, an encoder module is used to perform paragraph interactions, and graph convolutional networks are applied to enhance the interaction between paragraphs with dependencies. After that, a deep differential amplifier is applied to amplify the difference between paragraphs, and a weighted loss function is used to balance the attention of the categories. In addition, text augmentation and adversarial training are also employed to enhance the focus on minorities. Finally, ensemble learning is used to further improve the model's performance. The experimental results on MCDTB 2.0 demonstrate that our model FPRME outperforms the state-of-the-art models.
With the rapid development of deep learning technology, deep counterfeiting has also quietly ascended to the stage of history, and voice disguise fraud, as a new type of attack means, has laid a huge hidden danger for...
详细信息
The emerging QR Code Extended Storage technology has the potential to greatly advance information storage and retrieval. This research examines the current state of QR code technology, its limitations, and the opportu...
详细信息
The rapid advancement of smart home technologies necessitates efficient human activity recognition (HAR) systems while ensuring user privacy. This research presents a novel architecture that integrates deep learning a...
详细信息
this paper examines using data augmentation strategies in the ensemble, getting to know medical photo segmentation with transfer learning. Various transfer-gaining knowledge of techniques, namely pretrained models, un...
详细信息
Cloud computing is one of the current research areas in computer science. Recently, Cloud is the buzz word used everywhere in IT industries;It introduced the notion of 'pay as you use' and revolutionized devel...
详细信息
ISBN:
(纸本)9783031686382;9783031686399
Cloud computing is one of the current research areas in computer science. Recently, Cloud is the buzz word used everywhere in IT industries;It introduced the notion of 'pay as you use' and revolutionized developments in IT. The rapid growth of modernized cloud computing leads to 24 x 7 accessing of e-resources from anywhere at any time. It offers storage as a service where users' data can be stored on a cloud which is managed by a third party who is called Cloud Service Provider (CSP). Since users' data are managed by a third party, it must be encrypted ensuring confidentiality and privacy of the data. There are different types of cryptographic algorithms used for cloud security;in this article, the algorithms and their security measures are discussed.
With the continuous development of digital economy and Internet of Things technology, it has brought huge benefits to people and generated massive amounts of data. As a result, users migrate their data to the cloud. C...
详细信息
暂无评论