检索结果-内蒙古大学图书馆

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zhengbo Wang Jian Liang Zilei Wang Tieniu Tan University of Science and Technology of China CRIPAC & MAIS Institute of Automation Chinese Academy of Sciences Nanjing University

Zero-shot learning (ZSL) aims to recognize unseen classes by generalizing the relation between visual features and semantic attributes learned from the seen classes. A recent paradigm called transductive zero-shot learning further leverages unlabeled unseen data during training and has obtained impressive results. These methods always synthesize unseen features from attributes through a generative adversarial network to mitigate the bias towards seen classes. However, they neglect the semantic information in the unlabeled unseen data and thus fail to generate high-fidelity attribute-consistent unseen features. To address this issue, we present a novel transductive ZSL method that produces semantic attributes of the unseen data and imposes them on the generative process. In particular, we first train an attribute decoder that learns the mapping from visual features to semantic attributes. Then, from the attribute decoder, we obtain pseudo-attributes of unlabeled data and integrate them into the generative model, which helps capture the detailed differences within unseen classes so as to synthesize more discriminative features. Experiments on five standard benchmarks show that our method yields state-of-the-art results for zero-shot learning.

关键词： Training visualization Semantics Signal processing Benchmark testing Generative adversarial networks data models

来源：评论

学校读者我要写书评

暂无评论

SUVR: A Search-Based Approach to Unsupervised Visual Representation Learning

SUVR: A Search-Based Approach to Unsupervised Visual Represe...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Yi-Zhan Xu Chih-Yao Chen Cheng-Te Li Institute of Data Science National Cheng Kung University Academia Sinica Institute of Information Science

Unsupervised learning has grown in popularity because of the difficulty of collecting annotated data and the development of modern frameworks that allow us to learn from unlabeled data. Existing studies, however, either disregard variations at different levels of similarity or only consider negative samples from one batch. We argue that image pairs should have varying degrees of similarity, and the negative samples should be allowed to be drawn from the entire dataset. In this work, we propose Search-based Unsupervised Visual Representation Learning (SUVR) to learn better image representations in an unsupervised manner. We first construct a graph from the image dataset by the similarity between images, and adopt the concept of graph traversal to explore positive samples. In the meantime, we make sure that negative samples can be drawn from the full dataset. Quantitative experiments on five benchmark image classification datasets demonstrate that SUVR can significantly outperform strong competing methods on unsupervised embedding learning. Qualitative experiments also show that SUVR can produce better representations in which similar images are clustered closer together than unrelated images in the latent space.

关键词： Representation learning visualization Signal processing Image representation Benchmark testing Linear programming Task analysis

来源：评论

学校读者我要写书评

暂无评论

Structuring Genre Performance for Future data Scientists

Structuring Genre Performance for Future Data Scientists

引用

International Conference on Professional Communication (IPCC)

作者： Michael Laudenbach Allison Hutchison Zhiyu Guo Danielle Xu Carnegie Mellon University Cornell University

In our panel’s presentations, we will discuss how our approaches to curriculum design in data science can help researchers and instructors name the types of writing skills they are asking students to display—and to perform—in varying instantiations throughout their academic careers, as well as later workplace contexts. This is especially relevant for data-driven writing in technical and professional settings, which we address in the teaching of our respective courses at two universities. The panelists will present two complementary studies that use Write & Audit, a text visualization tool that displays disciplinary genre choices for students. The presenters stress that Write & Audit is a non-evaluative revision tool designed for students to make more rhetorically informed choices in their technical writing. The course and workshops we’ve designed represent an “ inter actionist” model, where writing and content knowledge are intertwined. Additionally, panelists will share survey results from their respective studies which capture students’ sense of communicative self-efficacy and motivation. Overall, both studies show that our interventions positively affected students’ learning in several areas. Therefore, we believe communication advances data analysis that is core to problem-solving efforts in the data science field.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Virtual, Augmented, and Extended Reality Applied to science Communication: A Systematic Literature Review

引用

ieee Transactions on visualization and Computer Graphics 2025年 PP卷 PP页

作者： Romero-Luis, Juan Rubio-Tamayo, José Luis Sanchez-Acedo, Alberto Wuebben, Daniel Codesido-Linares, Valeri Madrid Spain Science for Change Barcelona Spain Universidad Rey Juan Carlos Spain Universidad Pontificia de Comillas Spain

Extended reality (XR)—which includes virtual reality (VR) and augmented reality (AR)—is becoming increasingly popular for sharing scientific knowledge. This research evaluates the state-of-the-art in XR for scientific communication. Our two-phase methodology began with a Systematic Literature Review, identifying 94 relevant articles and conference papers from the last decade (2013- 2023) sourced from the Web of science and SCOPUS databases. These publications show scholars and practitioners using XR to convey scientific findings, foster awareness, ignite interest, shape opinions, and enhance understanding. In the second phase, we applied data clustering and analysis. Our findings highlight a significant increase in XR studies over the last decade, with the XR technologies used for communication (N = 24), dissemination (N = 23), educational/training (N = 21), and decision-making (N = 10). Our results indicate the need to establish clearer guidelines for aligning science communication and to create more possibilities to publish peer-reviewed research in. © 1995-2012 ieee.

关键词： Publishing

来源：评论

学校读者我要写书评

暂无评论

visualization Analytics of Mao Zedong Research in 2022 Based on Mapping Knowledge Domain

Visualization Analytics of Mao Zedong Research in 2022 Based...

引用

ieee International Conference on Cloud Computing and Big data Analysis (ICCCBDA)

作者： Miao Li Jinhua Yang School of Marxism Huazhong University of Science and Technology Wuhan China

In recent years, Mao Zedong’s research has received much attention by lots of researchers of multidisciplinary and publications. Particularly, the bibliometric analyses of Mao Zedong Research could induce the trend and hotspots for the inheritance and development of Mao Zedong Thought with the centenary of the founding the Communist Party of China (CPC). According to the visual and cluster analysis technology, this study analyses the development status, research distribution and future trends of Mao Zedong Research in the 2022. Firstly, a total number of 1204 published paper indexed by CNKI in 2022 are investigated. Then, according to the Mapping Knowledge Domain (MKD) approaches, the visualization of keywords, research organization and co-authorship analyses are presented based on the VOSviewer, which stimulates insightful findings. These cluster analysis results show that the Spiritual pedigree of the CPC, Party history education, fighting spirit, and national security thought of Mao Zedong are research hotspots, which could devote to the Party history education and the Partys theoretical innovation.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Machine Learning based Platform for Remote Management of Heart Failure Patients

A Machine Learning based Platform for Remote Management of H...

引用

ieee International Conference on Intelligent Computer Communication and Processing (ICCP)

作者： Rareș-Ioan Man Ionuț Anghel Computer Science Department Technical University of Cluj-Napoca Cluj-Napoca Romania

Heart failure is a growing concern due to its high incidence nowadays, also representing a major cause of morbidity and mortality worldwide. In this paper we propose a web-based platform that incorporates both the clinical data prediction aspect and the continuous monitoring of the heart health. We implement multiple machine learning models that can support the doctors in the process of classification between a healthy and unhealthy situation. The platform benefits from an ETL (Extract, Transform, Load) sub-system that processes biometric data from smart wearables and displays it in customizable dashboards for a more illustrative visualization. The prediction service integrates three Machine Learning (ML) techniques, namely Logistic Regression, Naïve Bayes Classifier, and a custom Artificial Neural Network responsible for making classifications on the monitored data. The results illustrate that the proposed solution’s usage for performing remote monitoring and heart health assessment is feasible, obtaining promising accuracies with the aid of a public heart failure dataset (best accuracy of 88.5%).

关键词：

来源：评论

学校读者我要写书评

暂无评论

Simulation Tool for Learning Action-Oriented Process Mining

Simulation Tool for Learning Action-Oriented Process Mining

引用

ieee Education Engineering (EDUCON)

作者： Clemens Drieschner Andreas Heckl Helmut Krcmar Department of Computer Science Technical University of Munich Garching Germany

Process mining enables companies to reduce costs by identifying bottlenecks, improving processes, or automation potentials, for example. Digital business process data, which can be extracted from information systems and afterward arranged in so-called event logs, allows the visualization and analysis of business processes. Several teaching materials for process mining are offered to learners and teacher. However, the existing data for learning and teaching process mining is static and does not change with specific user interactions, such as the release of invoices in other information systems. This work contributes to a more realistic teaching approach by providing a simulation tool to continuously create business process data reacting to user actions. First, the application scope of process mining projects in the industry is determined to collect use cases of action-oriented process mining from practice. Then, a prototypical implementation is presented that allows the continuous simulation of business process data that allows specific user interventions. A didactic framework is described to apply the simulation tool in the teaching of process mining. Finally, the simulation tool is discussed and evaluated by four experts from practice and academia.

关键词： Industries Education Prototypes Software data models data mining Environmental management

来源：评论

学校读者我要写书评

暂无评论

Deep Learning for Microstructure Segmentation and Defect Detection in Additive Manufacturing Systems

Deep Learning for Microstructure Segmentation and Defect Det...

引用

ieee/ACM Int'l Conference on & Int'l Conference on Cyber, Physical and Social Computing (CPSCom) Green Computing and Communications (GreenCom)

作者： Zhaochen Gu Venkata Mani Krishna Karri Shashank Sharma Hang Tran Aishwarya Manjunath Donger Chen Song Fu Narendra B. Dahotre Department of Computer Science and Engineering University of North Texas Denton Texas USA Department of Material Science and Engineering University of North Texas Denton Texas USA

Microstructure analysis plays a crucial role in additive manufacturing (AM) processes as it provides significant insights into material properties and print quality. Detecting melt pool boundaries and porosity defects within microstructure samples is a complex task due to inherent image processing and data analysis challenges. To address this, we propose a deep learning-based approach that employs state-of-the-art deep learning models with convolutional neural network architectures (e.g., U-Net and FPN). This approach enables automatic segmentation and detection of melt pools and porosity in AM microstructure images. visualization results demonstrate significant potential in accurately identifying microstructures using limited data. A comparative evaluation of various deep learning network architectures indicates that U-Net paired with EfficientNet b7 is better suited for melt pool segmentation, and the FPN with the DenseNet 201 backbone attains the highest accuracy for porosity. This work and the generated results demonstrate great promise in enhancing AM quality control through deep learning-powered microstructure analysis.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multimodal Facial Action unit Detection with Physiological Signals

Multimodal Facial Action unit Detection with Physiological S...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zhihua Li Lijun Yin Department of Computer Science Binghamton University NY USA

Most of the existing facial action unit detection models seek to improve detection accuracy by utilizing multiple visual modalities, including 3D geometry, thermal, and depth images. However, the potential usage of heterogeneous physiological modalities (e.g., heart rate and blood pressure) for AU detection is not considered in current works. Meanwhile, it’s challenging to fully utilize the hidden emotion-correlated physiological signals. In this paper, we propose deep networks to extract temporal features from periodic and non-periodic time-series signals and design an informativeness-based feature fusion module to handle the signal noise. Then, we utilize spatial-temporal visual representations to infer the physiological embeddings, allowing absent physiological data during testing. Experiments show that our multimodal framework achieves state-of-the-art performances on two widely used datasets: MMSE and BP4D.

关键词： Gold visualization Solid modeling Three-dimensional displays Signal processing Feature extraction Physiology

来源：评论

学校读者我要写书评

暂无评论

Training Recurrent Neural Network-Based Model to Predict COVID-19 Patient Risk for PASC using Pytorch_EHR – Hands-on Tutorial on N3C

Training Recurrent Neural Network-Based Model to Predict COV...

引用

ieee International Conference on Healthcare Informatics (ICHI)

作者： Laila Rasmy Ziqian Xie Degui Zhi School of Biomedical Informatics University of Texas Health Science Center Houston TX USA

Pytorch_EHR is a codebase enabling fast prototyping of deep learning-based predictive models using electronic health records structured data. Rather than a collection of vertical pipelines implementing methods from papers claiming state-of the-art results, Pytorch_EHR offers efficient implementations of data flow, padding, embeddings, choices of popular recurrent neural networks (RNN) cells and layers structures, prediction heads, and also predictions explanation and visualization. This tutorial will provide participants with a clear understanding of deep learning theoretical concepts behind Pytorch_EHR different components, as well as hands-on experience in building an explainable RNN-based model to solve a real-world clinical problem on a cloud-based platform hosting the National COVID Cohort Collaborative (N3C) data. URL: https://***/ZhiGroup/pytorch_ehr

关键词：

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：