相关参考文献
注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article
Mathematics
Rehan Raza et al.
Summary: This paper focuses on a specific type of melanoma, proposing a novel method for classification using stacked ensemble approach. Experimental results confirm the effectiveness of the proposed method, achieving an accuracy of 97.93%.
Article
Chemistry, Multidisciplinary
Andreea Bianca Popescu et al.
Summary: Deep learning algorithms have shown great potential in healthcare applications, but privacy concerns arise when the data needs to leave the healthcare facility. To address this, we propose an image obfuscation algorithm that protects the content of medical images and enables DL model training on secured data. The algorithm successfully protects the images without significant computational overhead.
APPLIED SCIENCES-BASEL
(2022)
Article
Engineering, Civil
Yijing Li et al.
Summary: This study introduces federated learning into autonomous driving to protect vehicular privacy, considering both honest-but-curious MEC servers and malicious vehicles, and proposing different privacy-preserving schemes based on traceable identity and anonymous identity. Simulation results show significant reductions in training loss and improvements in accuracy while maintaining effective privacy under the threat of dishonest MEC servers and vehicles.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Computer Science, Information Systems
Vishesh Kumar Tanwar et al.
Summary: This paper investigates privacy-preserving image recognition model for encrypted data over cloud, addressing security concerns by proposing a block-based image encryption scheme to protect visual information. The proposed method is proved to be secure through various cryptographic attacks, and experiments demonstrate that SecureDL successfully overcomes storage and computational overheads associated with fully-homomorphic and multi-party computation based secure recognition schemes.
JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION
(2022)
Article
Medicine, General & Internal
Muhammad Attique Khan et al.
Summary: This study proposes an automated framework for segmentation and classification of gastrointestinal diseases based on deep saliency maps and Bayesian optimal deep learning feature selection. It improves the original images through preprocessing and segments the infected regions, then fine-tunes a pre-trained model and extracts features using transfer learning. A hybrid optimization algorithm is used to select the best features, and an extreme learning machine classifier is employed for classification. The experimental results show improved accuracy compared to other methods.
Article
Computer Science, Information Systems
Jamuna S. Murthy et al.
Summary: This paper introduces a real-time object detection framework for advanced driver assistant systems (ADAS), which improves the speed of object detection by implementing the YOLOv5 algorithm. The paper compares other state-of-the-art object detectors and shows that YOLOv5 is faster and more accurate. The framework is used to build a mobile application called ObjectDetect to assist drivers in making better decisions on the road.
WIRELESS COMMUNICATIONS & MOBILE COMPUTING
(2022)
Article
Environmental Sciences
Wadii Boulila et al.
Summary: Deep learning has become an essential tool in remote sensing research, but it also brings challenges to data privacy and security. Privacy-preserving deep learning techniques can provide a solution. This study proposes a hybrid approach for object classification in high-resolution satellite images, using encryption methods to protect data privacy.
Article
Engineering, Multidisciplinary
Mujeeb Ur Rehman et al.
Summary: This study proposes a novel privacy-preserving non-invasive cancer detection method using Deep Learning. By encrypting clinical data and employing techniques such as Convolutional Neural Network model, high accuracy cancer detection is achieved.
IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING
(2022)
Review
Engineering, Multidisciplinary
Uddagiri Sirisha et al.
Summary: In our day-to-day life, synchronizing vision and language aspects plays a crucial role in solving various real-time challenges. Image captioning, which aims to provide syntactically and semantically correct visual descriptions, has been explored in various domains. This research article examines and analyzes different image captioning models across domains, and determines that LSTM performs best in multiple domains.
COGENT ENGINEERING
(2022)
Article
Computer Science, Information Systems
Wangsheng Fang et al.
Summary: The variable-speed elastic collision lion swarm algorithm (VELSO) was proposed by enhancing the flexibility and learning strategies of lioness search, increasing population diversity and individual quality, ultimately improving the algorithm's ability to obtain the optimal solution.
Proceedings Paper
Computer Science, Artificial Intelligence
Wadii Boulila et al.
Summary: In this study, a privacy-preserving deep learning approach is proposed for secure classification of Chest X-ray images. Experimental results demonstrate that the proposed method achieves high classification accuracy while protecting data privacy.
2022 2ND INTERNATIONAL CONFERENCE OF SMART SYSTEMS AND EMERGING TECHNOLOGIES (SMARTTECH 2022)
(2022)
Article
Computer Science, Information Systems
Qi-Xian Huang et al.
Summary: The need for cloud servers for training complex deep neural network (DNN) models is increasing. However, cloud servers are considered semi-honest and privacy is a concern. Previous studies have proposed learnable image encryption schemes but there is still room for improvement. We proposed an enhanced version of a learnable image encryption scheme that can train a successful DNN model while preserving the privacy of training images.
Article
Computer Science, Artificial Intelligence
J. Ashok Kumar et al.
Summary: Toxicity identification is a serious issue in online communities, and an automatic system like MCBiGRU is proposed for detecting toxic comments. Experimental results show that the MCBiGRU model outperforms in terms of multilabel metrics.
Article
Environmental Sciences
Munirah Alkhelaiwi et al.
Summary: This research proposes an efficient approach that utilizes privacy-preserving deep learning techniques to address privacy concerns regarding satellite image data, with the proposed encryption scheme allowing for processing of data without exposing the underlying information. Experiments demonstrate that CNN-based models retain data utility while maintaining data privacy.
Review
Computer Science, Information Systems
Raghida El Saj et al.
Summary: Privacy-preserving deep neural networks are crucial for protecting personal and sensitive data, and various methods have been developed to address the privacy protection issues using encrypted data in neural networks. Some of these methods have shown the ability to hide information, maintain high classification accuracy, and resist attacks, suggesting their potential application in other domains.
Proceedings Paper
Computer Science, Artificial Intelligence
S. Phani Praveen et al.
Summary: Cloud computing is an emerging technology with various services, including medical images storage. Security is crucial in cloud storage, especially for privacy-sensitive medical imaging data. The introduction of the Advanced Cipher-text Algorithm (ACTA) provides extra security tools and shows promising performance in experimental studies.
2021 IEEE INTERNATIONAL CONFERENCE ON IMAGING SYSTEMS AND TECHNIQUES (IST)
(2021)
Proceedings Paper
Computer Science, Hardware & Architecture
Wenbo He et al.
Summary: The paper proposes CryptoEyes to tackle the challenges of privacy-preserving classification over encrypted images. By sharing a secret sequence between the service provider and the image owner, CryptoEyes allows for improved classification accuracy and better privacy preservation performance. Experimental results demonstrate the superiority of CryptoEyes over existing state of the arts in terms of classification accuracy over encrypted images and privacy preservation.
IEEE CONFERENCE ON COMPUTER COMMUNICATIONS (IEEE INFOCOM 2021)
(2021)
Article
Computer Science, Artificial Intelligence
Georgios Kaissis et al.
Summary: PriMIA is a free, open-source software framework for privacy-preserving medical image analysis, performing well in instance testing and preventing data disclosure.
NATURE MACHINE INTELLIGENCE
(2021)
Article
Computer Science, Information Systems
Hiroki Ito et al.
Summary: The proposed image transformation network protects the visual information of plain images, maintains high classification accuracy, and is robust against attacks, without the need for managing security keys.
Article
Computer Science, Artificial Intelligence
Ali Pashaei et al.
JOURNAL OF REAL-TIME IMAGE PROCESSING
(2020)
Article
Computer Science, Information Systems
Lei Song et al.
Article
Computer Science, Theory & Methods
Kazuaki Nakamura et al.
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
(2019)
Article
Computer Science, Theory & Methods
Yanqi Zhao et al.
JOURNAL OF PARALLEL AND DISTRIBUTED COMPUTING
(2019)
Article
Computer Science, Information Systems
Meng Shen et al.
IEEE INTERNET OF THINGS JOURNAL
(2019)
Proceedings Paper
Computer Science, Information Systems
Hao Chen et al.
PROCEEDINGS OF THE 2019 ACM SIGSAC CONFERENCE ON COMPUTER AND COMMUNICATIONS SECURITY (CCS'19)
(2019)
Article
Computer Science, Interdisciplinary Applications
Milad Salem et al.