检索结果-内蒙古大学图书馆

Unsupervised Color Segmentation with Reconstructed Spatial Weighted Gaussian Mixture Model and Random Color Histogram

引用

computers, Materials & Continua 2024年第3期78卷 3323-3348页

作者： Umer Sadiq Khan Zhen Liu Fang Xu Muhib Ullah Khan Lerui Chen Touseef Ahmed Khan Muhammad Kashif Khattak Yuquan Zhang School of Computer and Information Science Hubei Engineering UniversityXiaogan432000China Institute for AI Industrial Technology Research Hubei Engineering UniversityXiaogan432000China Key Lab of Mountain Hazards and Surface Processes Institute of Mountain Hazards and EnvironmentChinese Academy of SciencesChengdu610041China University of Chinese Academy of Sciences Beijing100049China College of Aviation Zhongyuan University of TechnologyZhengzhou451191China Research Center of Digital Mountain and Remote Sensing Application Institute of Mountain Hazards and EnvironmentChinese Academy of SciencesChengdu610041China School of Computer Science University of PoonchAzad Jammu Kashmir12350Pakistan School of Computer Science and Information Engineering Hubei UniversityWuhan430062China

Image classification and unsupervised image segmentation can be achieved using the Gaussian mixture *** the Gaussian mixture model enhances the flexibility of image segmentation,it does not reflect spatial information and is sensitive to the segmentation *** this study,we first present an efficient algorithm that incorporates spatial information into the Gaussian mixture model(GMM)without parameter *** proposed model highlights the residual region with considerable information and constructs color ***,we incorporate the content-based color saliency as spatial information in the Gaussian mixture *** segmentation is performed by clustering each pixel into an appropriate component according to the expectation maximization and maximum ***,the random color histogram assigns a unique color to each cluster and creates an attractive color by default for segmentation.A random color histogram serves as an effective tool for data visualization and is instrumental in the creation of generative art,facilitating both analytical and aesthetic *** experiments,we have used the Berkeley segmentation dataset BSDS-500 and Microsoft Research in Cambridge *** the study,the proposed model showcases notable advancements in unsupervised image segmentation,with probabilistic rand index(PRI)values reaching 0.80,BDE scores as low as 12.25 and 12.02,compactness variations at 0.59 and 0.7,and variation of information(VI)reduced to 2.0 and 1.49 for the BSDS-500 and MSRC datasets,respectively,outperforming current leading-edge methods and yielding more precise segmentations.

关键词： Unsupervised segmentation color saliency spatial weighted GMM random color histogram

来源：评论

学校读者我要写书评

暂无评论

Ground-Glass Lung Nodules Recognition Based on CatBoost Feature Selection and Stacking Ensemble Learning

引用

Journal of Shanghai Jiaotong University (science) 2024年 1-10页

作者： Miao, Jun Chang, Yiru Chen, Chen Zhang, Maoxuan Liu, Yan Qi, Honggang Guo, Zhijun Xu, Qian School of Computer Science Beijing Information Science and Technology University Beijing100101 China Pathological Information Engineering Technology Center Jinan Supercomputing Technology Research Institute Jinan250100 China School of Computer Science and Technology University of the Chinese Academy of Sciences Beijing100049 China Department of Radiology Tianjin Kanghui Hospital Tianjin300385 China Department of Radiology Huabei Petroleum General Hospital Hebei Renqiu062550 China

Aimed at the issue of high feature dimensionality, excessive data redundancy, and low recognition accuracy of using single classifiers on ground-glass lung nodule recognition, a recognition method based on CatBoost feature selection and Stacking ensemble learning was proposed. First, the method uses a feature selection algorithm to filter important features and remove features with less impact, achieving the effect of data dimensionality reduction. Second, random forests classifier, decision trees, K-nearest neighbor classifier, and light gradient boosting machine were used as base classifiers, and support vector machine was used as meta classifier to fuse and construct the ensemble learning model. This measure increases the accuracy of the classification model while maintaining the diversity of the base classifiers. The experimental results show that the recognition accuracy of the proposed method reaches 94.375%. Compared to the random forest algorithm with the best performance among single classifiers, the accuracy of the proposed method is increased by 1.875%. Compared to the recent deep learning methods (ResNet+GBM+Attention and MVCSNet) on ground-glass pulmonary nodule recognition, the proposed method’s performance is also better or comparative. Experiments show that the proposed model can effectively select features and make recognition on ground-glass pulmonary nodules. © Shanghai Jiao Tong University 2024.

关键词： Feature Selection

来源：评论

学校读者我要写书评

暂无评论

Unidirectional Negative Refraction at an Exceptional Point of Acoustic P T-Symmetric Systems

引用

Chinese Physics Letters 2023年第12期40卷 56-61页

作者：刘宸蓝君顾仲明祝捷 Institute of Acoustics School of Physics Science and EngineeringTongji UniversityShanghai 200092China State Key Laboratory of Acoustics Chinese Academy of SciencesBeijing 100190China College of Computer Science and Technology Nanjing Tech UniversityNanjing 211800China

We demonstrate a method to realize unidirectional negative refraction in an acoustic parity-time(P T)-symmetric system, which is composed of a pair of metasurfaces sandwiching an air gap. The pair of metasurfaces possesses loss and gain modulations. The unidirectional negative refraction, which is strictly limited to the case of incident wave imposing on the loss end of the metasurface, is demonstrated at the exception point(EP) in this P T-symmetric system, while the incidence from the other side leads to strong reflection. Based on rigorous calculations, we explicitly show the underlying mechanism of this model to achieve unidirectional wave scatterings around the EP in the parametric space. In addition, the perfect imaging of a point source in the three-dimensional space, as a signature of negative refraction, is simulated to provide a verification of our work. We envision that this work may sharpen the understanding of P T-symmetric structures and inspire more acoustic functional devices.

关键词： refraction unidirectional symmetric

来源：评论

学校读者我要写书评

暂无评论

Malicious Traffic Detection with Noise Labels Based on Cross-Modal Consistency

IEEE Networking Letters

引用

IEEE Networking Letters 2024年第2期6卷 148-151页

作者： Yuan, Qingjun Niu, Weina Wang, Yongjuan Gou, Gaopeng Lu, Bin Henan Key Laboratory of Network Cryptography Technology Zhengzhou450001 China Institute for Cyber Security School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu611731 China Shenzhen Institute for Advanced Study University of Electronic Science and Technology of China Shenzhen518110 China Institute of Information Engineering Chinese Academy of Sciences Beijing100093 China

To train robust malicious traffic identification models under noisy labeled datasets, a number of learning with noise labels approaches have been introduced, among which parallel training methods have been proved to be effective. Parallel training methods tend to select samples with disagreement to mitigate the risk of self-control degradation. However, it also introduces noisy knowledge into training. In this letter, we try to avoid introducing noisy knowledge by enhancing the consistency of the representations of parallel networks. Meanwhile, the two networks are heterogeneous and introduce information from different modalities, thus mitigating the risk of self-control degradation from multiple perspectives. © 2019 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Developing a Convolutional Neural Model for Segmentation of Medical Images 1

Developing a Convolutional Neural Model for Segmentation of ...

引用

1st IEEE International Conference on Emerging Research in Computational science, ICERCS 2023

作者： Arunachalam, M. Kumar, P. Mohan Selvam, L. Kumar, S. Senthil Arunkumar, B. Sri Krishna College of Engineering and Technology Department of Information Technology Coimbatore India Hindusthan Institute of Technology & Sciences Department of Computer Science and Engineering Chennai India Karpagam Academy of Higher Education Department of Computer Science and Engineering Coimbatore India Sethu Institute of Technology Department of Computer Science and Engineering Kariapatti India

ISBN: (纸本)9798350359756

Medical image segmentation is a new biomedical image processing method that has made a significant contribution to sustainable health care. It has now become a major study area in the realm of computer vision research. Medical image processing based on deep convolutional neural networks has become a research hotspot due to the fast development of deep learning techniques. An additional cascade component is included to the architecture in this article to better replicate spatial closure tags' direct dependences. It is also utilised for post-segmentation processing, resolving the conflict between classification performance and network intensity and number of mixing times in a typical convolutional net. Numerous state-of-the-art deep recognition models, and a multi-atlas segmentation methodology, are outperformed by our methodology. © 2023 IEEE.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Memory-enhanced text style transfer with dynamic style learning and calibration

引用

science China(Information sciences) 2024年第4期67卷 181-196页

作者： Fuqiang LIN Yiping SONG Zhiliang TIAN Wangqun CHEN Diwen DONG Bo LIU College of Computer Science and Technology National University of Defense Technology Department of Computer Science and Engineering The Hong Kong University of Science and Technology Strategic Assessments and Consultation Institute Academy of Military Sciences

Text style transfer aims to rephrase a sentence to match the desired style while retaining the original content. As a controllable text generation task, mainstream approaches use content-independent style embedding as control variables to guide stylistic generation. Nonetheless, stylistic properties are contextsensitive even under the same style. For example, “delicious” and “helpful” convey positive sentiments,although they are more likely to describe food and people, respectively. Therefore, desired style signals must vary with the content. To this end, we propose a memory-enhanced transfer method, which learns fine-grained style representation concerning content to assist transfer. Rather than employing static style embedding or latent variables, our method abstracts linguistic characteristics from training corpora and memorizes subdivided content with the corresponding style representations. The style signal is dynamically retrieved from memory using the content as a query, providing a more expressive and flexible latent style space. To address the imbalance between quantity and quality in different content, we further introduce a calibration method to augment memory construction by modeling the relationship between candidate *** results obtained using three benchmark datasets confirm the superior performance of our model compared to competitive approaches. The evaluation metrics and case study also indicate that our model can generate diverse stylistic phrases matching context.

关键词： style transfer memory-enhanced method text generation deep learning text representation

来源：评论

学校读者我要写书评

暂无评论

An Efficient Depth Estimation Technique for Visual Perception

An Efficient Depth Estimation Technique for Visual Perceptio...

引用

2024 IEEE International Conference on Recent Advances in science and engineering technology, ICRASET 2024

作者： Deepa Shubham, A. Shetty, Bhuvan Aravind Dhanush Kumar, Archana Praveen Department of Information Science and Engineering Karnataka Nitte India Manipal Institute of Technology Manipal Academy of Higher Education Department of Computer Science and Engineering Karnataka Manipal576104 India

ISBN: (纸本)9798350388602

In the realm of computer vision and 3D reconstruction, the accurate conversion of images into depth maps is crucial. This paper focuses on depth estimation techniques to assess their accuracy and reliability in generating depth maps from 2D images. Subsequently, the depth maps were transformed into 3D representations using Open3D in Python, enabling the visualization and reconstruction of three-dimensional scenes. Despite challenges such as noise and artifacts, the reconstruction process yielded promising results, highlighting the potential of image-based 3D reconstruction techniques. This study contributes to the advancement of computer vision technologies by providing insights into improving depth estimation accuracy and enhancing the fidelity of reconstructed 3D models. Additionally, the successful reconstruction of 3D scenes underscores the practical utility of image-based 3D reconstruction methodologies in various domains, including robotics, augmented reality, and autonomous driving. © 2024 IEEE.

关键词： Machine vision

来源：评论

学校读者我要写书评

暂无评论

Intelligent Recognition of Optical Modulation Formats in Dynamic Communication Networks 1

Intelligent Recognition of Optical Modulation Formats in Dyn...

引用

1st International Conference on Software, Systems and Information technology, SSITCON 2024

作者： Gnanakumar, D. Anusooya, V. Muthupradeepa, R. Gobalakrishnan, S. Kalasalingam Academy of Research and Education Department of Computer Science and Engineering Krishnankoil India Srm Institute of Science and Technology Department of Electronics and Communication Engineering Kattankulathur India Thandalam India

ISBN: (纸本)9798350352931

Machine learning-based optical modulation format recognition is essential for dynamic optical networks. Convolutional Neural Networks (CNNs) can analyze signal space diagrams directly from raw data. Specifically, CNNs excel in image classification due to their ability to extract features and learn autonomously. In this study, signal space diagrams for various modulation formats (BPSK, QPSK, 8-QAM, 16-QAM, 32-QAM, and 64-QAM) are generated using an electrical constellation visualizer in Optisystem software, with the Optical Signal-to-Noise Ratio (OSNR) varying from 10 to 40 dB and the noise figure ranging from 5 to 25 dB. The proposed VGG-16 CNN architecture achieves 100% accuracy in modulation format recognition (MFR) for all six formats, utilizing minimal training data and fewer epochs. This approach can significantly enhance next-generation dynamic optical communication networks by enabling automatic detection and recognition of modulation formats. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

LPCC Based Music Genrefication Using Hybrid Computational Model 7

LPCC Based Music Genrefication Using Hybrid Computational Mo...

引用

2023 7th International Conference On Computing, Communication, Control And Automation, ICCUBEA 2023

作者： Sul, Astha Sarda, Harshit Billa, Aishwarya Mishra, Tusar Kanti School of Computer Science and Engineering Vellore Institute of Technology Vellore India Manipal Institute of Technology Bengaluru Manipal Academy of Higher Education Dept. of Computer Science and Engineering Manipal India

ISBN: (纸本)9798350304268

This work deals with spontaneous music genrefication through computational models which in the recent times has been gaining importance rapidly. Through these hybrid computational models implemented users get an enhanced level of satisfaction when their choice of music files genre with least latency is gained. The paper includes the use of LPCC attributes to obtain the features of the music files along with robust classification models such as neural networks. For this purpose, one thousand audio files are taken as the sample set. The input audio file is pre-processed and suitable LPCC features are computed and stored into final vector. The pool of vectors of numerous audio samples are finally utilized to train a neural model. The proposed work trains the model for ten very distinct categories of music such as hip-hop, jazz, metal, pop, blues, classical, country, disco, reggae, and rock. Further, a comparison is also made between all the classifiers such as SVM, ANN and random forest (RF). Comparatively the best accuracy rate of 85.53\% has been achieved for the proposed work that validates its effectiveness. © 2023 IEEE.

关键词： ANN gtzan dataset LPCC Music Genre

来源：评论

学校读者我要写书评

暂无评论

Poster Abstract: On the Accuracy and Robustness of Large Language Models in Chinese Industrial Scenarios 23

Poster Abstract: On the Accuracy and Robustness of Large Lan...

引用

23rd ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2024

作者： Li, Zongjie Qiu, Wenying Ma, Pingchuan Li, Yichen Li, You He, Sijia Jiang, Baozheng Wang, Shuai Gu, Weixi Hong Kong University of Science and Technology Department of Computer Science and Engineering Hong Kong Hong Kong China Academy of Industrial Internet Beijing China

ISBN: (纸本)9798350362015

Recent studies have demonstrated that large language models (LLMs) exhibit exceptional performance across various natural language processing tasks, rivaling or even exceeding human competencies in certain areas [1] - [5]. Typically, LLMs undergo pre-training on extensive text corpora, usually using billions of tokens to develop a foundational model. To better align LLMs with human preferences and directives or to fulfill specific application needs, methods such as supervised fine-tuning (SFT), reinforcement learning from human feedback (RLHF), and direct preference optimization (DPO) have been introduced and demonstrated to be effective. These advancements facilitate more intuitive and efficient human-AI interactions. However, the substantial resource requirements throughout the training process pose challenges for individual users and smaller organizations. © 2024 IEEE.

关键词： Reinforcement 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：