检索结果-内蒙古大学图书馆

24th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2024

作者： Li, Ying Shan, Chun Liu, Zhen Liao, Shuyan Beijing Key Laboratory of Software Security Engineering Technology Beijing Institute of Technology Beijing China School of Computer Science and Technology Beijing Institute of Technology Beijing China Computer School Beijing Information Science and Technology University Beijing China School of Cyberspace Science and Technology Beijing Institute of Technology Beijing China

ISBN: (纸本)9789819615476

Deep learning(DL) systems exhibit multiple behavioral characteristics such as correctness, robustness, and fairness. Ensuring that these behavioral characteristics function properly is crucial for maintaining the accuracy of DL systems’ outputs. As a specialized form of software, DL systems’ security testing techniques have increasingly become a focus of research in quality assurance. We analyze and organize the testing techniques for DL systems based on an investigation of the current state of the art both domestically and internationally. This paper categorizes existing approaches as component-oriented and attribute-oriented methods, providing a detailed review based on this classification. Additionally, we forecast the future development of testing techniques for DL systems. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

A survey on deep learning for polyp segmentation: techniques, challenges and future trends

引用

Visual Intelligence 2025年第1期3卷 1-20页

作者： Mei, Jiaxin Zhou, Tao Huang, Kaiwen Zhang, Yizhe Zhou, Yi Wu, Ye Fu, Huazhu PCA Lab Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education Nanjing China School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing China School of Computer Science and Engineering Southeast University Nanjing China Institute of High Performance Computing A*STAR Singapore Singapore

Early detection and assessment of polyps play a crucial role in the prevention and treatment of colorectal cancer (CRC). Polyp segmentation provides an effective solution to assist clinicians in accurately locating and segmenting polyp regions. In the past, people often relied on manually extracted lower-level features such as color, texture, and shape, which often had problems capturing global context and lacked robustness to complex scenarios. With the advent of deep learning, more and more medical image segmentation algorithms based on deep learning networks have emerged, making significant progress in the field. This paper provides a comprehensive review of polyp segmentation algorithms. We first review some traditional algorithms based on manually extracted features and deep segmentation algorithms, and then describe benchmark datasets related to the topic. Specifically, we carry out a comprehensive evaluation of recent deep learning models and results based on polyp size, taking into account the focus of research topics and differences in network structures. Finally, we discuss the challenges of polyp segmentation and future trends in the field. © The Author(s) 2025.

关键词： Comprehensive evaluation Deep learning Medical imaging Polyp segmentation

来源：评论

学校读者我要写书评

暂无评论

DDA: Distillation-Driven Acceleration of the Reverse Diffusion Process for Stochastic Multi-Ship Trajectory Prediction

DDA: Distillation-Driven Acceleration of the Reverse Diffusi...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Ma, Kun Han, Qilong Yao, Jingzheng Wu, Changmao Zhang, Yuntao College of Computer Science and Technology Harbin Engineering University Harbin China Yantai Research Institute Harbin Engineering University Yantai China Institute of Software Chinese Academy of Sciences Beijing China College of Shipbuilding Engineering Harbin Engineering University Harbin China

ISBN: (纸本)9798350368741

Modeling stochastic multi-ship trajectories is vital for maritime safety and interaction efficiency. Recent researches show that diffusion models excel in trajectory prediction, surpassing GANs and VAEs in generation quality, diversity and stability. However, their slow sampling speed remains a major limitation, as producing high-quality trajectories typically requires hundreds of denoising steps. We introduce DDA, a novel method that accelerates multi-ship trajectory generation by distilling the reverse diffusion process, progressively reducing sampling steps by half while minimizing quality loss. We use CVAE-based encoder to map multimodal inputs into state embeddings in the latent space, and use distillation diffusion in the latent space to more quickly and better represent multi-ship trajectories. The diffusion model uses Transformer-based core, and we incorporate SO(2) invariance and equivariance to enhance model representation. Validation on real-world AIS datasets shows that the student model retains high-quality trajectory generation while sampling speed is approximately 30 times faster. © 2025 IEEE.

关键词： Accelerating Sampling AIS Diffusion Model Progressive Distillation VAE

来源：评论

学校读者我要写书评

暂无评论

Evaluating the Impact of Pulse Oximetry Bias in Machine Learning Under Counterfactual Thinking 3rd

Evaluating the Impact of Pulse Oximetry Bias in Machine L...

引用

3rd International Workshop on Applications of Medical Artificial Intelligence, AMAI 2024 held in conjunction with the 27th International Conference on Medical Image Computing and computer Assisted Intervention, MICCAI 2024

作者： Martins, Inês Matos, João Gonçalves, Tiago Celi, Leo A. Wong, An-Kwok Ian Cardoso, Jaime S. Faculty of Engineering University of Porto Porto Portugal Institute for Systems and Computer Engineering Technology and Science Porto Portugal Duke University Durham United States Massachusetts Institute of Technology Cambridge United States

ISBN: (纸本)9783031820069

Algorithmic bias in healthcare mirrors existing data biases. However, the factors driving unfairness are not always known. Medical devices capture significant amounts of data but are prone to errors;for instance, pulse oximeters overestimate the arterial oxygen saturation of darker-skinned individuals, leading to worse outcomes. The impact of this bias in machine learning (ML) models remains unclear. This study addresses the technical challenges of quantifying the impact of medical device bias in downstream ML. Our experiments compare a "perfect world", without pulse oximetry bias, using SaO2 (blood-gas), to the "actual world", with biased measurements, using SpO2 (pulse oximetry). Under this counterfactual design, two models are trained with identical data, features, and settings, except for the method of measuring oxygen saturation: models using SaO2 are a "control" and models using SpO2 a "treatment". The blood-gas oximetry linked dataset was a suitable test-bed, containing 163,396 nearly-simultaneous SpO2 - SaO2 paired measurements, aligned with a wide array of clinical features and outcomes. We studied three classification tasks: in-hospital mortality, respiratory SOFA score in the next 24 h, and SOFA score increase by two points. Models using SaO2 instead of SpO2 generally showed better performance. Patients with overestimation of O2 by pulse oximetry of ≥ 3% had significant decreases in mortality prediction recall, from 0.63 to 0.59, P © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Oximeters

来源：评论

学校读者我要写书评

暂无评论

Algorithms for the Collaborative Delivery Problem with Monitored Constraints 19th

Algorithms for the Collaborative Delivery Problem with Mo...

引用

19th International Conference and Workshops on Algorithms and Computation, WALCOM 2025

作者： Blank, Lotte Huynh, Kien C. Luo, Kelin Naredla, Anurag Murty Institute of Computer Science University of Bonn Bonn Germany Communications and Transport Systems ITN Linköping University Linköping Sweden Department of Computer Science and Engineering University at Buffalo GetzvilleNY United States

ISBN: (纸本)9789819628445

We provide efficient algorithms to solve package delivery problems in which a sequence of drones work together to ‘optimally’ deliver a package from a source s to a target t. The package may be transferred from one drone to another only on a given line on which secure package transfers are ensured (we refer to this as monitored constraints). We allow the source and target to lie outside of this line. The drones have different starting locations, speeds, and rates of fuel consumption. Two notions of optimality are studied—fuel-efficient delivery (minimize total fuel consumption in a successful delivery), and fastest delivery (minimize total time for delivery). © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Drones

来源：评论

学校读者我要写书评

暂无评论

Machine Learning in Higher Education using RF and SVM for Pedagogical Enhancement 3

Machine Learning in Higher Education using RF and SVM for Pe...

引用

3rd International Conference on Integrated Circuits and Communication systems, ICICACS 2025

作者： Kunwar, Fateh Bahadur Srivastava, Satyajee Dhivya, R. Gulati, Monika Anand, Taruna Raj, D. Antony Arul Geetanjali Institute of Technical Studies Department of Computer Science and Engineering Udaipur India School of Engineering and Technology Manav Rachna International Institute of Research and Studies Department of Computer Science and Engineering Faridabad India M.Kumarasamy College of Engineering Department of Information Technology Karur India Ambala India University Department of Humanities and Social Sciences Dehradun India Psg College of Arts & Science Department of Software Systems Coimbatore India

ISBN: (纸本)9798331508456

AI's revolutionary potential in higher education is examined in this proposal, including how it could revolutionize teaching, learning, administration, and research. Adaptive learning platforms, intelligent tutoring systems, and virtual learning assistants may change how students learn by improving engagement, providing quick feedback, and enabling tailored learning. The suggested process includes preprocessing, feature extraction, and model training. Preprocessing involves de-anonymizing, tokenizing, and removing punctuation, spaces, and numbers before analysis. The linear transformation method PCA extracts features. Training the systems uses the RF -SVM approach, which combines the advantages of RF and SVM. The suggested RF -SVM model beats two RF and SVM models with an average accuracy of 90.39 *** improves resource use, strategic decision-making, and research capacity in university management. The RF-SVM model's success in academia highlights the practicality of AI-driven solutions. © 2025 IEEE.

关键词： Teaching

来源：评论

学校读者我要写书评

暂无评论

Multilevel Discrete Cosine Transform and CNN-Based Image Steganography 3

Multilevel Discrete Cosine Transform and CNN-Based Image Ste...

引用

3rd International Conference on Intelligent systems, Advanced Computing, and Communication, ISACC 2025

作者： Subburaj, Brindha Sai, Aditya Shantanu, Shantanu Won, Daehan Vellore Institute of Technology School of Computer Science and Engineering Chennai India Binghamton University Department of Systems Science and Industrial Engineering United States

ISBN: (纸本)9798331523893

This paper investigates the growing role of steganography in cybersecurity and presents a hybrid implementation called Multi-level Discrete Cosine Convolution (MDCC) that applies a Multi-level Discrete Cosine Transform (MDCT) and a Convolutional Neural Network (CNN). Steganography provides some advantages compared to cryptography, mainly its invisibility, but there are weaknesses in information capacity and resistance to attacks in traditional techniques. The paper suggests solving these difficulties with a new steganography method using deep learning. The proposed MDCC method provides an opportunity to encrypt large amounts of data in realistic images that are beyond human visual perception. The system is trained on the ImageNet dataset by using a loss function to balance the quality of the cover image and the information it conceals within it. The MDCC model proposed yields accurate results for all metrics since PSNR, SSIM, and bits per pixel come out to be 80.13dB, 0.9918, and 12.5 bpp, respectively. © 2025 IEEE.

关键词： Convolutional Neural Networks Deep Learning Discrete Cosine Transform Embedding Steganography

来源：评论

学校读者我要写书评

暂无评论

IPL Data Analysis by Web Scraping and Using PowerBI

IPL Data Analysis by Web Scraping and Using PowerBI

引用

IT Innovation and Knowledge Discovery (ITIKD), International Conference on

作者： Mogith P Mothiswar T B G Mohammad Abubak J. Jagannathan N. Balaganesh Software Engineering Vellore Institute of Technology Vellore India School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore India

ISBN: (数字)9798350355468

ISBN: (纸本)9798350355475

Cricket fans throughout the world avidly follow the Indian Premier League (IPL), a competition of great fame. The IPL data is examined in this using Exploratory Data Analysis (EDA) approaches to find hidden trends, patterns, and insights. Key findings are summarized in EDA, a critical stage of data processing that is frequently complemented with visual aids. Through EDA, the goal is to investigate several factors, including team performance, player data, match results, and venue effects. Data transformation, cleaning, and visualization will all be part of producing insightful findings. Win percentages, player averages, and run distributions are examples of key performance indicators (KPIs) that will be looked at. The analysis's findings will give teams, athletes, and supporters important new information.

关键词： Visualization Technological innovation Fans Data analysis Key performance indicator Data visualization Market research Knowledge discovery Cleaning

来源：评论

学校读者我要写书评

暂无评论

Enhanced Student Stress Prediction in Higher Education Using a TriBranch CNN Framework 3

Enhanced Student Stress Prediction in Higher Education Using...

引用

3rd International Conference on Integrated Circuits and Communication systems, ICICACS 2025

作者： Kunwar, Fateh Bahadur Srivastava, Satyajee Dhivya, R. Gulati, Monika Amudha, K. Raj, D. Antony Arul Geetanjali Institute of Technical Studies Department of Computer Science and Engineering Udaipur India School of Engineering and Technology Manav Rachna International Institute of Research and Studies Department of Computer Science and Engineering Faridabad India M.Kumarasamy College of Engineering Department of Information Technology Karur India Mullana Ambala India Kongunadu College of Engineering and Technology Department of Electronics and Communication Engineering Trichy India Psg College of Arts & Science Department of Software Systems Coimbatore India

ISBN: (纸本)9798331508456

Students' stress levels have a significant impact on their academic performance in higher education, thus it's important to identify signs of stress early on and intervene to improve academic performance. Machine learning has evolved into a powerful tool for stress prediction, allowing for proactive support of at-risk youngsters. In order to better anticipate college students' stress levels, this study presents an enhanced TriBranch CNN framework. In order to provide the best possible performance, the technique includes data preparation, feature extraction, and model training. The dimensionality and redundancy are reduced using Principal Component Analysis (PCA). After being trained and tested against other models, the Fully Conditional Temporal Bayesian Network (Fully CTBN) outperformed them all with an accuracy of 93.29%, beating both traditional CNN and FCN approaches. These findings demonstrate that the Fully CTBN approach is useful for identifying students who may be struggling academically and providing schools with the information they need to implement data-driven interventions that boost student well-being and performance. © 2025 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Deepfake Detection using XAI based Deep Fusion Models 2

Deepfake Detection using XAI based Deep Fusion Models

引用

2nd International Conference on Computational Intelligence, Communication Technology and Networking, CICTN 2025

作者： Anitha, A. Saju, Lis Maria Kamaraj, Balakrishnan School of Computer Science Engineering and Information Systems Vellore Institute of Technology Tamil Nadu Vellore India Madurai Medical College Tamilnadu Madurai India

ISBN: (纸本)9798331530389

Deepfake technology has rapidly evolved, posing a serious threat to the authenticity of digital media and contributing to the spread of misinformation. The manipulation of media content raises significant concerns across sectors like politics, entertainment, and social media, undermining public trust in digital information. In response, this research proposes an advanced deepfake detection model that integrates Explainable Artificial Intelligence (XAI) techniques within a hybrid deep learning architecture. The proposed model combines ResNet50 and Vision Transformer (ViT) to capture spatial and contextual features effectively. While exploring additional methods such as facial landmark detection and frequency domain analysis, the final model focuses on the ResNet50-ViT combination for its superior performance to fusion model. To ensure transparency, the model employs XAI techniques such as LIME (Local Interpretable Model-agnostic Explanations) and SHAP (SHapley Additive explanations), offering interpretability by revealing the key factors that influence its predictions. This dual focus on high detection accuracy and interpretability ensures that the model not only detects deepfakes effectively but also builds trust in automated decision-making processes. By leveraging cutting-edge deep learning and XAI methods, this framework offers a more reliable, transparent, and adaptable solution for preserving digital media authenticity across various applications, ultimately contributing to combating misinformation and fostering trust in digital content. © 2025 IEEE.

关键词： Contrastive Learning

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：