检索结果-内蒙古大学图书馆

SteriCNN: Cloud native stego content sterilization framework

JOURNAL OF INFORMATION SECURITY AND APPLICATIONS 2024年 87卷

作者： Banerjee, Abhisek Ganguly, Sreeparna Mukherjee, Imon Ganguly, Nabanita Indian Inst Informat Technol Kalyani 741235 India Maulana Abul Kalam Azad Univ Technol Kolkata 700064 India

Modern robust steganography-based cyber attacks often bypass intrinsic cloud security measures, and contemporary steganalysis methods struggle to address these covert threats due to recent advancements in deep learning (DL)-based steganography techniques. Existing steganography removal methods are constrained by trade-offs involving high processing times, poor quality of sanitized images, and insufficient removal of steganographic content. This paper introduces SteriCNN, a lightweight deep residual neural network model designed for steganography removal. SteriCNN effectively eliminates embedded steganographic information while preserving the visual integrity of the sanitized images. We employ a series of convolutional blocks with three residual connections for feature extraction, feature learning, feature attention, and image reconstruction from the residue. The proposed model utilizes the correlation of channel features to achieve a faster learning rate, and by varying the dilation rate in convolutional blocks, the model achieves wider receptive fields, enabling it to cover larger areas of the input image at each layer. SteriCNN is targeted for blind image sterilization for real-time use cases due to its low training and prediction time costs. Our study shows impressive results for both traditional and deep learning-based stego vulnerabilities, with approximately 90% of steganograms eliminated while maintaining an average PSNR value of 46 dB and an SSIM of 0.99 when tested with popular steganography methods.

关键词： Cloud security Steganography Steganalysis image stego content removal Residual neural network

来源：评论

学校读者我要写书评

暂无评论

Reducing false positive rate with the help of scene change indicator in deep learning based real-time face recognition systems

引用

MULtimeDIA TOOLS AND APPLICATIONS 2023年第30期82卷 47517-47536页

作者： Kutlugun, Mehmet Ali Sirin, Yahya Istanbul Sabahattin Zaim Univ Comp Sci & Engn Istanbul Turkiye

In face recognition systems, light direction, reflection, and emotional and physical changes on the face are some of the main factors that make recognition difficult. Researchers continue to work on deep learning-based algorithms to overcome these difficulties. It is essential to develop models that will work with high accuracy and reduce the computational cost, especially in real-time face recognition systems. deep metric learning algorithms called representative learning are frequently preferred in this field. However, in addition to the extraction of outstanding representative features, the appropriate classification of these feature vectors is also an essential factor affecting the performance. The Scene Change Indicator (SCI) in this study is proposed to reduce or eliminate false recognition rates in sliding windows with a deep metric learning model. This model detects the blocks where the scene does not change and tries to identify the comparison threshold value used in the classifier stage with a new value more precisely. Increasing the sensitivity ratio across the unchanging scene blocks allows for fewer comparisons among the samples in the database. The model proposed in the experimental study reached 99.25% accuracy and 99.28% F-1 score values compared to the original deep metric learning model. Experimental results show that even if there are differences in facial images of the same person in unchanging scenes, misrecognition can be minimized because the sample area being compared is narrowed.

关键词： Classification deep metric learning Illumination and pose changes image processing real-time face recognition

来源：评论

学校读者我要写书评

暂无评论

deep learning in metallography: adapting a grain size determination model to develop the new comparison charts for ISO 643

引用

Practical Metallography 2025年第3期62卷 176-194页

作者： Wu, Zhaobo Li, Jikang Zhang, Jianing Stücklin, Stephan Central Iron and Steel Research Institute Company Limited Beijing100081 China Swiss Steel Group Emmenbrücke Switzerland

The determination of average grain size is an important component in the microstructural characterization of metallic materials. The grain size is usually determined using the intercept and comparison methods, but there are problems with the relatively time-consuming intercept method and the low accuracy of the comparison method. In this paper, a new model is proposed to realize the task of automatic grain size grading using a combination of traditional image processing and deep learning. deep learning semantic segmentation and image classification are used to achieve the grading of different quality images. The model is capable of extracting and complementing grain boundaries and calculating the pixel area of the grains. The accuracy of the model was tested on metallographic phases of pure iron and austenitic stainless steel. Using this model, grain size comparison charts were produced for ISO 643 based on real metallographic photographs. © 2025 Walter de Gruyter GmbH, Berlin/Boston.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Optimizing image retrieval by leveraging YCbCr colour space quadtree segmentation and deep learning models for enhanced accuracy and efficiency

引用

IMAGING SCIENCE JOURNAL 2025年

作者： Damahe, Lalit Yenurkar, Ganesh Mane, Sulakshana Ramotra, Atul Kumar Jarbais, Goldi Nyangaresi, Vincent O. Yeshwantrao Chavan Coll Engn Dept Comp Sci & Engn Nagpur India Yeshwantrao Chavan Coll Engn Dept Comp Technol Nagpur India Bharati Vidyapeeth Coll Engn Dept Comp Engn Mumbai India ACE Engn Coll Dept CSE AI & ML Hyderabad Telangana India Jaram Oginga Odinga Univ Sci & Technol Dept Comp Sci & Engn Bondo Kenya SIMATS Saveetha Sch Engn Dept Appl Econ Chennai Tamilnadu India

In today's rapidly evolving digital landscape, the demand for multimedia applications is surging, driven by significant advancements in computer and storage technologies that enable efficient compression and storage of visual data in large-scale databases. However, challenges such as inaccuracy, inefficiency, and suboptimal precision and recall in image retrieval systems necessitate the development of faster and more reliable techniques for searching and retrieving images. Traditional retrieval systems often rely on RGB colour spaces, which may inadequately represent critical image information. In response, we propose a content-based image retrieval (CBIR) system that integrates advanced techniques such as quadtree segmentation alongside modern lightweight deep learning models, specifically MobileNet and EfficientNet, to enhance precision and recall. Our comparative experiments reveal that these deep learning models significantly outperform traditional methods, including SVM classifiers combined with feature extraction techniques such as Histogram of Oriented Gradients (HOG), Scale-Invariant Feature Transform (SIFT), and Speeded-Up Robust Features (SURF). Notably, MobileNet and EfficientNet achieved F1-scores of 0.87 and 0.89, respectively, with enhanced processing efficiencies that resulted in feature extraction times reduced to 20 ms and classification times down to 8 ms. This translates to rapid image retrieval times as low as 35 ms, highlighting the superior performance of modern deep learning models in enhancing both retrieval accuracy and efficiency for large-scale image databases, making them ideal for real-time applications.

关键词： CBIR HSV YCbCr quadtree segmentation query image multimedia applications MobileNet EfficientNet

来源：评论

学校读者我要写书评

暂无评论

Rocknet: lightweight network for real-time segmentation of Martian rocks

引用

JOURNAL OF real-time image processing 2025年第1期22卷 1-11页

作者： Wei, Pengfei Sun, Zezhou Tian, He Jilin Univ Sch Mech & Aerosp Engn Changchun 130025 Peoples R China Beijing Inst Spacecraft Syst Engn Beijing 100094 Peoples R China Beijing Key Lab Intelligent Space Robot Syst Techn Beijing 100094 Peoples R China

Rock segmentation on the Martian is particularly critical for rover navigation, obstacle avoidance, and scientific target detection. We propose a lightweight network for real-time semantic segmentation of Martian rocks (RockNet). First, we propose the cross-dimension channel attention (CDCA) model to replace traditional downsample and upsample operation, which gives more weight to the channels with more useful information by adjusting the weight of each channel. Second, we modify the short-term dense concatenate model, we adopt dilated convolution to learn the feature with a larger receptive field, and through the skip connection structure, the degradation of the network can be reduced. Finally, we propose a feature fusion module (FFM) to fully fuse different levels of features. With only 0.86M parameters, our model gets 82.37% mIoU and 105.7 FPS running speed on the dataset of TWMARS.

关键词： Rock segmentation deep learning Lightweight network Encoder-decoder

来源：评论

学校读者我要写书评

暂无评论

ShaderNN: A lightweight and efficient inference engine for real-time applications on mobile GPUs

引用

NEUROCOMPUTING 2025年 611卷 128628页

作者： Xie, Jing Yan, Yuzhong Saxena, Abhishek Qiu, Qiang Chen, Jiangong Sun, Hongyu Chen, Rong Bhattacharyya, Shuvra S. OPPO Seattle Res Ctr 10940 NE 33rd Pl 202 Bellevue WA 98004 USA Univ Maryland Coll Pk Dept Elect & Comp Engn 8223 Paint Branch Dr College Pk MD 20740 USA Univ Maryland Baltimore Dept Diagnost Radiol & Nucl Med 22 South Greene St First floor Baltimore MD 21201 USA Univ Maryland Coll Pk Inst Adv Comp Studies 8125 Paint Branch Dr College Pk MD 20740 USA

Inference using deep neural networks on mobile devices has been an active area of research in recent years. The design of a deep learning inference framework targeted for mobile devices needs to consider various factors, such as the limited computational capacity of the devices, low power budget, varied memory access methods, and I/O bus bandwidth governed by the underlying processor's architecture. Furthermore, integrating an inference framework with time-sensitive applications - such as games and video-based software to perform tasks like ray tracing denoising and video processing - introduces the need to minimize data movement between processors and increase data locality in the target processor. In this paper, we propose Shader Neural Network (ShaderNN), an OpenGL-based, fast, and power-efficient inference framework designed for mobile devices to address these challenges. Our contributions include the following: (1) the texture-based input/output provides an efficient, zero-copy integration with real-time graphics pipelines or image processing applications, thereby saving expensive data transfers between CPU and GPU, which are unavoidable in most existing inference engines;(2) we are the first to leverage fragment shaders based on the OpenGL backend in neural network inference operators, which has an advantage in deploying parametrically small neural network models;(3) a hybrid implementation of the compute shader and fragment shader is proposed that enables layer-level shader selection to boost performance;and (4) we utilize OpenGL features - such as normalization, interpolation and texture padding - to improve performance. Experiments illustrate the favorable performance of ShaderNN over other popular on-device deep learning frameworks such as TensorFlow-Lite on the latest mobile devices powered by Qualcomm and MediaTek chips. A case study further demonstrates the usability and integration of the ShaderNN framework with a media processing Android application s

关键词： deep learning Inference Shader

来源：评论

学校读者我要写书评

暂无评论

DREAM-PCD: deep Reconstruction and Enhancement of mmWave Radar Pointcloud

引用

IEEE TRANSACTIONS ON image processing 2024年 33卷 6774-6789页

作者： Geng, Ruixu Li, Yadong Zhang, Dongheng Wu, Jincheng Gao, Yating Hu, Yang Chen, Yan Univ Sci & Technol China Sch Cyber Sci & Technol Hefei 230026 Peoples R China Univ Washington Dept Elect & Comp Engn Seattle WA 98195 USA Univ Sci & Technol China Sch Informat Sci & Technol Hefei 230026 Peoples R China

Millimeter-wave (mmWave) radar pointcloud offers attractive potential for 3D sensing, thanks to its robustness in challenging conditions such as smoke and low illumination. However, existing methods failed to simultaneously address the three main challenges in mmWave radar pointcloud reconstruction: specular information lost, low angular resolution, and severe interference. In this paper, we propose DREAM-PCD, a novel framework specifically designed for real-time 3D environment sensing that combines signal processing and deep learning methods into three well-designed components to tackle all three challenges: Non-Coherent Accumulation for dense points, Synthetic Aperture Accumulation for improved angular resolution, and real-Denoise Multiframe network for interference removal. By leveraging causal multiple viewpoints accumulation and the "real-denoise" mechanism, DREAM-PCD significantly enhances the generalization performance and real-time capability. We also introduce RadarEyes, the largest mmWave indoor dataset with over 1,000,000 frames, featuring a unique design incorporating two orthogonal single-chip radars, Lidar, and camera, enriching dataset diversity and applications. Experimental results demonstrate that DREAM-PCD surpasses existing methods in reconstruction quality, and exhibits superior generalization and real-time capabilities, enabling high-quality real-time reconstruction of radar pointcloud under various parameters and scenarios. We believe that DREAM-PCD, along with the RadarEyes dataset, will significantly advance mmWave radar perception in future real-world applications.

关键词： Radar Millimeter wave communication Radar imaging Three-dimensional displays Laser radar Interference Estimation image reconstruction Synthetic aperture radar Sensors mmWave imaging super-resolution pointcloud

来源：评论

学校读者我要写书评

暂无评论

Inter - intra observer variability using deep learning and traditional image processing for breast cancer 13

Inter - intra observer variability using deep learning and t...

引用

13th International Conference on Developments in eSystems Engineering (DeSE)

作者： Almazroa, Ahmed Alsomaie, Barrak Alluhaydan, Najd Alhaidary, Amal Fahim, Mohammed Abdul, Wadood Alawad, Mohammed Khan, Altaf Alenezi, Ebtihal Almotairi, Taghreed Alyahya, AlJowharah Alfulayj, Raghad Althobaiti, Manar bin Maythir, Jawaher King Saud bin Abdulaziz Univ Hlth Sci King Abdullah Int Med Res Ctr Med Imaging Dept Riyadh Saudi Arabia King Abdul Aziz Med City Med Imaging Dept Riyadh Saudi Arabia King Saud Univ Coll Comp & Informat Sci Dept Comp Engn Riyadh Saudi Arabia King Saud bin Abdulaziz Univ Hlth Sci King Abdullah Int Med Res Ctr Biostat & Bioinformat Dept Riyadh Saudi Arabia King Saud bin Abdulaziz Univ Hlth Sci Radiol Sci Dept Riyadh Saudi Arabia

ISBN: (纸本)9781665422383

Breast cancer is one of the most common life-threatening diseases that affects women globally. Saudi Arabia is also one of the countries that suffer from a serious number of this disease among women. In terms of diagnosis modalities, a mammogram is the first line for detecting breast cancer. In addition, breast cancer can be screened by real-time ultrasound images, which are of relatively less quality and have more impact (noninvasive) images. Therefore, the purpose of this study is to develop image enhancement techniques using deep learning and image processing techniques. The main goal is to improve the ultrasound images in order to help radiologists screen the disease more accurately. For this study, ninety female patients of ages between 15 - 77 years are considered. These patients were already diagnosed using ultrasound with breast lesions. The images are visually graded and evaluated by two trained radiologists, both pre- and post- enhancement. In particular, two parameters were considered;1) BI-RAD categories, 2) Breast cancer classification. The agreement between radiologists and post-enhancement was assessed using simple kappa and weighted kappa statistics. Moreover, sensitivity and specificity are also calculated.

关键词： Breast cancer image processing deep learning Computer-aided diagnosis systems

来源：评论

学校读者我要写书评

暂无评论

Advancing Underwater Vision: A Survey of deep learning Models for Underwater Object Recognition and Tracking

引用

IEEE ACCESS 2025年 13卷 17830-17867页

作者： Elmezain, Mahmoud Saad Saoud, Lyes Sultan, Atif Heshmat, Mohamed Seneviratne, Lakmal Hussain, Irfan Khalifa Univ Khalifa Univ Ctr Autonomous Robot Syst Abu Dhabi U Arab Emirates

Underwater computer vision plays a vital role in ocean research, enabling autonomous navigation, infrastructure inspections, and marine life monitoring. However, the underwater environment presents unique challenges, including color distortion, limited visibility, and dynamic light conditions, which hinder the performance of traditional image processing methods. Recent advancements in deep learning (DL) have demonstrated remarkable success in overcoming these challenges by enabling robust feature extraction, image enhancement, and object recognition. This review provides a comprehensive analysis of cutting-edge deep learning architectures designed for underwater object detection, segmentation, and tracking. State-of-the-art (SOTA) models, including AGW-YOLOv8, Feature-Adaptive FPN, and Dual-SAM, have shown substantial improvements in addressing occlusions, camouflaging, and small underwater object detection. For tracking tasks, transformer-based models like SiamFCA and FishTrack leverage hierarchical attention mechanisms and convolutional neural networks (CNNs) to achieve high accuracy and robustness in dynamic underwater environments. Beyond optical imaging, this review explores alternative modalities such as sonar, hyperspectral imaging, and event-based vision, which provide complementary data to enhance underwater vision systems. These approaches improve performance under challenging conditions, enabling richer and more informative scene interpretation. Promising future directions are also discussed, emphasizing the need for domain adaptation techniques to improve generalizability, lightweight architectures for real-time performance, and multi-modal data fusion to enhance interpretability and robustness. By critically evaluating current methodologies and highlighting gaps, this review provides insights for advancing underwater computer vision systems to support ocean exploration, ecological conservation, and disaster management.

关键词： Reviews Object recognition Computer vision Computational modeling Imaging image segmentation deep learning Surveys Optical imaging Oceans Underwater computer vision deep learning underwater robotics ocean research underwater image enhancement object tracking object detection

来源：评论

学校读者我要写书评

暂无评论

Enhancing vehicle fault diagnosis through multi-view sound analysis: integrating scalograms and spectrograms in a deep learning framework

引用

SIGNAL image AND VIDEO processing 2025年第1期19卷 1-19页

作者： Akbalik, Ferit Yildiz, Abdulnasir Ertugrul, Omer Faruk Zan, Hasan Batman Univ Social Sci Vocat Sch Batman Turkiye Dicle Univ Dept Elect & Elect Engn Diyarbakir Turkiye Batman Univ Dept Elect & Elect Engn Batman Turkiye Mardin Artuklu Univ Dept Comp Engn Mardin Turkiye

This study presents a comprehensive framework for vehicle fault diagnosis using engine sound signals, leveraging deep learning models and a multi-view approach. Traditional methods for vehicle fault diagnosis often rely on the expertise of mechanics or diagnostic tools, which can be costly, time-consuming, and may not always provide accurate results. To address these limitations, we propose CarFaultNet, a multi-view model that processes both scalograms and spectrograms simultaneously to capture complementary information from these time-frequency representations. Our approach incorporates transfer learning with pretrained convolutional neural networks, including AlexNet, GoogLeNet, ShuffleNet, SqueezeNet, and MobileNet v2, as well as CarFaultNet, which combines two MobileNet networks. The results demonstrate that CarFaultNet outperforms traditional machine learning methods and single-view deep learning models, achieving a precision of 95.32%, recall of 94.83%, F1-score of 94.99%, and accuracy of 95.00%. Class activation mapping visualizations provide valuable insights into the model's decision-making process, highlighting the regions of the input images that are most influential for the classification of different vehicle faults. By leveraging a large, diverse dataset encompassing various vehicle models and real-world operating conditions, our approach addresses the drawbacks of previous studies and demonstrates the potential of deep learning for practical and effective vehicle fault diagnosis.

关键词： Vehicle fault detection Pretrained models MobileNet Engine sound Scalogram Spectrogram, Class activation mapping

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：