检索结果-内蒙古大学图书馆

2nd International Conference on Machine Learning and Autonomous Systems, ICMLAS 2025

作者： Nandini, N. Vanajakshi, P. Vidyarani, H.J. Vani, B.A. Vinod Kumar, K.P. Shwethasree, R. Dr. Ambedkar Institute of Technology Department of Computer Science and Engineering Karnataka Bangalore India Vivekananda Institute of Technology Department of Information Science and Engineering Karnataka Bangalore India Sri Krishna Institute of Technology Department of Information Science and Engineering Karnataka Bangalore India

ISBN: (纸本)9798331505745

Tumors in the brains may be Benign or malignant but they are growths of cells in the brain. Aches, seizures, vomiting, dementia, and impaired vision are generic symptoms of the ailment. In early diagnosis, detection of the disease and determining the right treatment method for the patients MRI scans are important. There are a number of Deep Learning-based detection techniques available, however their accuracy and error rates may be low. This study uses a Random-Coupled Neural Network with Cat Hunting Optimization (RCNNet-CHOpt) to identify and classify brain tumor in order to get around these issues. Only one kind of BraTS dataset is used by the suggested approach: The filters to enhance image quality and reduce noise, the input images are pre-processed using the Adaptive and Propagated Mesh Filtering (APMF). This is followed by feature extraction and classification using Mellin Transform (M-Trans) with Random-Coupled Neural Network (R-CNN). When combined, these two networks can accurately classify the images and identify important elements. Finally, to increase accuracy, the weight parameter of R-CNN is modified using the Cat Hunting Optimization (CHO) technique. The proposed method has a 99.8% recall and a 99.9% accuracy rate, outperforming existing methods. Among the primary advantages of the RCNNet-CHOpt are decreased false-positive rates and enhanced precision in identifying intricate brain tumor formations. Hence, it is an especially useful tool for clinical use in the identification of brain tumor because most of these applications manage large volumes of medical data. © 2025 IEEE.

关键词： Neuroimaging

来源：评论

学校读者我要写书评

暂无评论

DSSM: Dual State Space Model For Human Motions Generation

DSSM: Dual State Space Model For Human Motions Generation

引用

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

作者： Liu, Yiming Zhao, Huan Liu, Yaqian Chen, Haijiao Li, Bo Ye, Guanghui Zhang, Zixing School of Computer Science and Electronic Engineering Hunan University Changsha410082 China Army Military Transportation University Tianjin Tianjin300161 China

ISBN: (纸本)9798350368741

Text-driven human motion generation has attracted considerable critical attention in recent years. The task requires generating movements that are diverse, natural, and comfortable in accordance with the text description. However, while generating the human motion, there is a significant gap in the amount of feature information contained in word-based text description modality and joint-based human motion modality. This results in an extreme imbalance of feature information in the latent space across the modalities, which seriously affects the effect of feature fusion. To alleviate the imbalance between modalities, we propose the Dual State Space Model (DSSM), which reconstruction the fused feature from coarse-to-fine. The DSSM contains two unit structures: the Masked State Space Model (MSSM) and the Hierarchical State Space Block (HSSB). At the same time, in order to make better use of timing information and reduce the computational complexity of the model, the DSSM is also the first method to introduce the state space model (SSM) into the text-driven motion sequence generation. We evaluated the DSSM on the HumanML3D and KITML benchmark datasets, and the experimental results show that our approach achieves state-of-the-art performance. The source code is available on GitHub at https://***/yimingliu123/DSSM. © 2025 IEEE.

关键词： Attention-free stateful model Motion Generation State space modal

来源：评论

学校读者我要写书评

暂无评论

Facial Action Unit Recognition with Micro-Action-Aware Transformer 2nd

Facial Action Unit Recognition with Micro-Action-Aware Tran...

引用

2nd CSIG Conference on Emotional Intelligence, CEI 2024

作者： Yuan, Yichen Cheng, Yifan Shao, Zhiwen Dang, Qianwen Chen, Rui Fu, Mingjian Jiang, Shengtian Li, Chunyu Ma, Lizhuang School of Computer Science and Technology China University of Mining and Technology Xuzhou China Mine Digitization Engineering Research Center of the Ministry of Education Xuzhou China Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai China

ISBN: (纸本)9789819650835

Facial action unit (AU) recognition is a challenging task, due to the subtlety of each AU and the correlations among AUs in global face. However, the learning of local-global features has not been thoroughly exploited in most of the existing methods. In this paper, we propose a novel micro-action-aware transformer to integrate local and global feature extractions, which effectively captures subtle AU details while maintaining the global relational modeling capacity of transformers. Besides, we jointly train facial AU recognition and facial landmark detection, in which the two correlated tasks contribute to each other and further facilitate the learning of local-global AU-related feature. Extensive experiments demonstrate that our approach achieves comparable performance to the state-of-the-art AU recognition methods on the challenging BP4D and GFT benchmarks, and works well for landmark detection. Particularly, our approach achieves average F1 score results of 63.3% and 55.8% on BP4D and GFT datasets, respectively. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： facial action unit recognition facial landmark detection Micro-action-aware transformer

来源：评论

学校读者我要写书评

暂无评论

Hyperspectral Images for Crop Prediction using Machine and Deep Learning Models 3

Hyperspectral Images for Crop Prediction using Machine and D...

引用

3rd International Conference on Electronics and Renewable Systems, ICEARS 2025

作者： Mahalakshmi, R. Kumutha, K. Sakthivanitha, M. Lakshmi, R. Bagavathi Sirajudeen, Mohamed Padmanabhan, Sankar Vels Institute of Science Technology & Advanced Studies Department of Computer Applications-PG Tamil Nadu Chennai India Vels Institute of Science Technology Advanced Studies Department of Computer Applications-PG Tamil Nadu Chennai India Vels Institute of Science Technology & Advanced Studies Department of Computer Application Tamil Nadu Chennai India Vels Institute of Science Technology & Advanced Studies Department of Information Technology Tamil Nadu Chennai India Nilgiri College of Arts and Science Pg and Research Department of Computer Science Tami Nadu The Nilgiris India Hindustan Institute of Technology and Science Department of Electronics and Communication Engineering Tamil Nadu Chennai India

ISBN: (纸本)9798331509675

Crop classification is essential for local and national governments to make informed agricultural decisions. Remote sensing technology has made it possible to employ high-resolution hyperspectral images (HSIs) for land cover classification for decades. Machine Learning (ML) and Deep Learning (DL) are becoming more popular for HSI categorization due to advances in processing capacity and technology. HSIs can accurately identify crop types due to their narrow and continuous spectral band reflection. This study compares traditional machine learning (ML) models like Support Vector Machines (SVM) and Random Forest (RF) to deep learning (DL) models like Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) for hyperspectral image classification (HSI). The HSI categorization was evaluated using performance criteria such as accuracy, precision, recall, and the F1-Score. CNN achieved the highest accuracy of 96%, demonstrating that it is the most reliable solution for complex HSI classification tasks because it successfully captures spatial and spectral data. © 2025 IEEE.

关键词： Classification Deep Learning(DL) Hyperspectral Images(HSI) Machine Learning(ML)

来源：评论

学校读者我要写书评

暂无评论

Decentralized Continuous Group Key Agreement for UAV Ad-Hoc Network 6th

Decentralized Continuous Group Key Agreement for UAV Ad-Hoc...

引用

6th International Conference on Machine Learning for Cyber Security, ML4CS 2024

作者： Hong, Shijia Zhou, Tianqi Yang, Huijie Shen, Mingdi Zheng, Wenying School of Information Science and Engineering Zhejiang Sci-Tech University Hangzhou310018 China School of Computer Science and Technology Zhejiang Sci-Tech University Hangzhou310018 China

ISBN: (纸本)9789819645657

With the rise of the Industrial Internet of Things (IIoT), Unmanned Aerial Vehicles (UAVs) have emerged as the preferred sensing tools for data collection. However, resource-constrained UAVs often face challenges in handling complex group key agreement processes, while key management in the flying ad hoc network (FANET) is also susceptible to single points of failure. To address this issue, this paper proposes a blockchain-based continuous group key agreement protocol for FANET. The blockchain is collectively maintained by UAVs and serves to record dynamic changes within the FANET. Additionally, it employs a ratchet tree to integrate a pseudo-random function generator with an updatable public key encryption scheme, effectively mitigating the risk of single points of failure, reducing communication overhead, and enhancing security and self-healing capabilities. Simulation experiments demonstrate that this protocol outperforms existing key agreement schemes in terms of reduced delay and improved security. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Blockchain

来源：评论

学校读者我要写书评

暂无评论

Multi-Mode Inflow Prediction for Urban Transportation Systems 14th

Multi-Mode Inflow Prediction for Urban Transportation System...

引用

14th International Conference on Frontier Computing, FC 2024

作者： Hung, Chih-Chieh Hung, Min-Hsien Chang, Jia-Wei Department of Management Information Systems National Chung Hsing University Taichung Taiwan Department of Computer Science and Information Engineering National Taichung University of Science and Technology Taichung Taiwan

ISBN: (纸本)9789819623945

Traffic congestion poses significant challenges to modern cities, leading to increased energy use, pollution, and long commute times. Optimizing public transit systems and encouraging their use is an effective solution. However, public transit integrates various modes like subways, buses, and shared bikes, requiring transfers between modes. Therefore, understanding the relationships between modes and accurately predicting passenger flows for each mode is crucial. Accurate predictions enable better capacity allocation, scheduling, and seamless transfers between modes, enhancing accessibility and convenience to attract more riders. Existing multi-modal prediction methods typically focus on station-to-region or region-to-region flows, overlooking station-to-station predictions that can better capture inter-modal relationships and provide fine-grained insights for traffic management. This study proposes a multi-modal framework for MRT and YouBike data that employs a spatial-temporal module to effectively model the spatiotemporal dependencies in the data. Evaluations on real-world datasets demonstrate the framework outperforms benchmark methods. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Traffic congestion

来源：评论

学校读者我要写书评

暂无评论

A novel approach to deriving adaboost classifier weights using squared loss function for overlapping speech detection

引用

Multimedia Tools and Applications 2025年 1-28页

作者： Asbai, Nassim Bounazou, Hadjer Zitouni, Sihem Department of Electronics Faculty of Sciences and Technology Mohamed El Bachir El Ibrahimi University Bordj Bou Arreridj City34030 Algeria Speech Com. & Signal Proc. Lab. Faculty of Electrical Engineering USTHB Bab Ezzouar City16111 Algeria Department of Computer Science Faculty of Mathematics and Computer Science Mohamed El Bachir El Ibrahimi University Bordj Bou Arreridj City34030 Algeria

In this paper, we address the complex problem of detecting overlapping speech segments, a key challenge in speech processing with applications in speaker diarization, automatic transcription, and multi-speaker recognition systems. Traditional approaches, often relying on exponential loss functions within the AdaBoost framework, struggle to maintain robustness in noisy or imbalanced data environments. To enhance detection accuracy, we propose a novel AdaBoost variant—SLFAdaBoost—utilizing a squared loss function specifically tailored for overlapping speech detection. This model demonstrates superior stability and convergence, significantly outperforming conventional methods. Experimental evaluation on the NIST 2005 corpus revealed that SLFAdaBoost achieves an accuracy range of 92.5% to 98.6%, with F1 scores between 95.4% and 99.2%, notably surpassing standard AdaBoost and Random Forest classifiers. Additionally, our model shows resilience in noisy conditions, maintaining precision and recall metrics at higher noise levels than other classifiers. These results underscore SLFAdaBoost's capacity to handle intricate data patterns, offering a more robust solution for real-world speech processing applications where overlapping segments are prevalent. This contribution provides an efficient, high-performing model, advancing the capabilities of ensemble methods in complex audio environments. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

MATRIX COMPLETION AND DECOMPOSITION IN PHASE-BOUNDED CONES

引用

SIAM Journal on Matrix Analysis and Applications 2025年第2期46卷 837-857页

作者： Zhang, Ding Ringh, Axel Li, Qiu Department of Electronic and Computer Engineering Hong Kong University of Science and Technology Clear Water Bay Kowloon Hong Kong Hong Kong Department of Mathematical Sciences Chalmers University of Technology the University of Gothenburg Gothenburg SE-412 96 Sweden School of Science and Engineering The Chinese Univ of Hong Kong Guangdong Shenzhen China

The problems of matrix completion and decomposition in the cone of positive semidefinite (PSD) matrices are well-understood problems, with many important applications in areas such as linear algebra, optimization, and control theory. This paper considers the completion and decomposition problems in a broader class of cones, namely, phase-bounded cones. We show that most of the main results from the PSD case carry over to the phase-bounded case. More precisely, this is done by first unveiling a duality between the completion and decomposition problems, using a dual-cone interpretation. Based on this, we then derive necessary and sufficient conditions for the phase-bounded completion and decomposition problems, and also characterize all phase-bounded completions of a completable partial matrix with a block banded pattern. Copyright © by SIAM.

关键词： Cones

来源：评论

学校读者我要写书评

暂无评论

Intelligent Traffic Management System using Multi-Agent Reinforcement Learning 3

Intelligent Traffic Management System using Multi-Agent Rein...

引用

3rd International Conference on Electronics and Renewable Systems, ICEARS 2025

作者： Shaheen, Dalaali Paulraj, Getzi Jeba Leelipushpam Jebadurai, Immanuel Johnraja Kirubakaran, Stewart Karunya Institute of Technology and Sciences Division of Computer Science and Engineering Coimbatore India Karunya Institute of Technology and Sciences Division of Cse Coimbatore India

ISBN: (纸本)9798331509675

This paper presents a decentralized multi-agent system for intelligent traffic management in urban environments, where each agent represents a traffic light controller at an intersection. The proposed system leverages a combination of Distributed W-Learning and Deep Q-Networks (DQN) to optimize traffic flow. Distributed W-Learning enables agents to prioritize decisions based on multiple performance policies, while DQN enhances their ability to handle complex state-action mappings through neural network approximations. By utilizing locally available traffic data, each agent adapts dynamically to diverse traffic conditions. The integration of these machine learning techniques ensures a fully decentralized and self-organizing approach, minimizing congestion, improving traffic efficiency, and addressing multiple objectives simultaneously. Simulation experiments are conducted using SUMO (Simulation of Urban MObility) to evaluate the system's performance in realistic traffic scenarios. This study highlights the potential of combining Distributed W-Learning and DQN in multi-agent reinforcement learning to revolutionize traffic management systems and achieve scalable, adaptive, and efficient urban traffic control. © 2025 IEEE.

关键词： Traffic congestion

来源：评论

学校读者我要写书评

暂无评论

Healthcare Fraud Detection Using an Integrated ML Approach with SMOTE

引用

Procedia computer science 2025年 258卷 800-810页

作者： Dipti Dash Mukesh Kumar Shaswati Patra Ajay Kumar Ankit Ganguly School of Computer Engineering Kalinga Institute of Industrial Technology-751024 Computer Science and Engineering Government Engineering College Kaimur School of Mechanical Engineering Kalinga Institute of Industrial Technology-751024

Healthcare fraud is a growing problem with significant financial ramifications for healthcare systems worldwide, demanding effective detection measures to preserve resources and assure the delivery of excellent treatment. This paper presents a novel approach to fraud detection that integrates Random Forest and K-Nearest Neighbors (KNN) algorithms with the Synthetic Minority Over-sampling Technique (SMOTE) to effectively address the challenges of identifying fraudulent healthcare claims. Our research tackles the common issue of class imbalance in training datasets, where legitimate transactions vastly outnumber fraudulent ones. By generating synthetic samples of fraudulent claims, SMOTE enhances the model’s learning capabilities, leading to improved detection accuracy. Our approach consisted of thorough data preparation, the creation of an integrated model, and a rigorous assessment using several performance measures including accuracy, precision, recall, and F1-score. The model exhibits exceptional performance in detecting fraud, achieving remarkable scores in accuracy 97.96%, precision 94.96%, F1-score 97.41%, and AUC 96.44%. This research ofers a promising solution to healthcare fraud by leveraging sophisticated methodologies, empowering healthcare providers and insurers to better safeguard their resources.

关键词： Machine Learning Synthetic Minority Over-sampling Technique (SMOTE) Class Imbalance Data Preprocessing Confusion Matrix

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：