In daily life,keyword spotting plays an important role in human-computer interaction.However,noise often interferes with the extraction of time-frequency information,and achieving both computational efficiency and rec...In daily life,keyword spotting plays an important role in human-computer interaction.However,noise often interferes with the extraction of time-frequency information,and achieving both computational efficiency and recognition accuracy on resource-constrained devices such as mobile terminals remains a major challenge.To address this,we propose a novel time-frequency dual-branch parallel residual network,which integrates a Dual-Branch Broadcast Residual module and a Time-Frequency Coordinate Attention module.The time-domain and frequency-domain branches are designed in parallel to independently extract temporal and spectral features,effectively avoiding the potential information loss caused by serial stacking,while enhancing information flow and multi-scale feature fusion.In terms of training strategy,a curriculum learning approach is introduced to progressively improve model robustness fromeasy to difficult tasks.Experimental results demonstrate that the proposed method consistently outperforms existing lightweight models under various signal-to-noise ratio(SNR)conditions,achieving superior far-field recognition performance on the Google Speech Commands V2 dataset.Notably,the model maintains stable performance even in low-SNR environments such as–10 dB,and generalizes well to unseen SNR conditions during training,validating its robustness to novel noise scenarios.Furthermore,the proposed model exhibits significantly fewer parameters,making it highly suitable for deployment on resource-limited devices.Overall,the model achieves a favorable balance between performance and parameter efficiency,demonstrating strong potential for practical applications.展开更多
In this paper, the specific application of key words Spotting used in the network monitoring is studied, and the keywords spotting is emphasized. The whole monitoring system is divided into two mod-ules： network moni...In this paper, the specific application of key words Spotting used in the network monitoring is studied, and the keywords spotting is emphasized. The whole monitoring system is divided into two mod-ules： network monitoring and keywords spotting. In the part of network monitoring, this paper adopts a method which is based on ARP spoofing technology to monitor the users＇ data, and to obtain the original audio streams. In the part of keywords spotting, the extraction methods of PLP （one of the main characteristic arameters） is studied, and improved feature parameters- PMCC are put forward. Meanwhile, in order to accurately detect syllable, the paper the double-threshold method with variance of frequency band method, and use the latter to carry out endpoint detection. Finally, keywords recognition module is built by HMM, and identification results are contrasted under Matlab environment. From the experiment results, a better solution for the application of key words recognition technology in network monitoring is found.展开更多
This paper focuses on developing a system that allows presentation authors to effectively retrieve presentation slides for reuse from a large volume of existing presentation materials. We assume episodic memories of t...This paper focuses on developing a system that allows presentation authors to effectively retrieve presentation slides for reuse from a large volume of existing presentation materials. We assume episodic memories of the authors can be used as contextual keywords in query expressions to efficiently dig out the expected slides for reuse rather than using only the part-of-slide-descriptions-based keyword queries. As a system, a new slide repository is proposed, composed of slide material collections, slide content data and pieces of information from authors＇ episodic memories related to each slide and presentation together with a slide retrieval application enabling authors to use the episodic memories as part of queries. The result of our experiment shows that the episodic memory-used queries can give more discoverability than the keyword-based queries. Additionally, an improvement model is discussed on the slide retrieval for further slide-finding efficiency by expanding the episodic memories model in the repository taking in the links with the author-and-slide-related data and events having been post on the private and social media sites.展开更多
The e-mail network is a type of social network. This study analyzes user behavior in e-mail subject participation in organizations by using social network analysis. First, the Enron dataset and the position-related in...The e-mail network is a type of social network. This study analyzes user behavior in e-mail subject participation in organizations by using social network analysis. First, the Enron dataset and the position-related information of an employee are introduced, and methods for deletion of false data are presented. Next, the three-layer model(User, Subject, Keyword) is proposed for analysis of user behavior. Then, the proposed keyword selection algorithm based on a greedy approach, and the influence and propagation of an e-mail subject are defined. Finally, the e-mail user behavior is analyzed for the Enron organization. This study has considerable significance in subject recommendation and character recognition.展开更多
Purpose:To reveal the research hotpots and relationship among three research hot topics in b iomedicine,namely CRISPR,iPS(induced Pluripotent Stem)cell and Synthetic biology.Design/methodology/approach:We set up their...Purpose:To reveal the research hotpots and relationship among three research hot topics in b iomedicine,namely CRISPR,iPS(induced Pluripotent Stem)cell and Synthetic biology.Design/methodology/approach:We set up their keyword co-occurrence networks with using three indicators and information visualization for metric analysis.Findings:The results reveal the main research hotspots in the three topics are different,but the overlapping keywords in the three topics indicate that they are mutually integrated and interacted each other.Research limitations:All analyses use keywords,without any other forms.Practical implications:We try to find the information distribution and structure of these three hot topics for revealing their research status and interactions,and for promoting biomedical developments.Originality/value:We chose the core keywords in three research hot topics in biomedicine by using h-index.展开更多
[目的]分析中医药治疗多发性硬化(multiple sclerosis,MS)的研究现状及热点,为其临床决策、后续研究及中西医结合诊疗指南构建提供参考。[方法]以中国知网、维普、万方、PubMed、Web of Science数据库为文献来源,检索2004年1月至2024年1...[目的]分析中医药治疗多发性硬化(multiple sclerosis,MS)的研究现状及热点,为其临床决策、后续研究及中西医结合诊疗指南构建提供参考。[方法]以中国知网、维普、万方、PubMed、Web of Science数据库为文献来源,检索2004年1月至2024年11月收录的相关文献。采用CiteSpace 6.3.R1软件对文献发表时间、作者、机构、关键词进行共现、聚类及突现等可视化分析,得出其研究现状及热点。[结果]共纳入文献168篇,中医药治疗MS领域的发文量近年无明显变化,而较高质量的临床研究占比增多;目前尚未形成稳定的核心作者群,研究机构集中在各大高校附属医院及中医药院校;高频和高中心性关键词包括辨证论治、名医经验、中西医结合、病因病机和中医证候。[结论]中医药治疗MS处于发展阶段,重点在病因病机及辨证论治研究,未来建议加强团队与机构合作,发展多中心网络,探索多学科交叉、多手段治疗,提高研究质量,为中医药治疗MS的临床决策提供高质量循证证据,以期完善治疗方案,构建中西医结合诊疗指南。展开更多
目的系统梳理药食同源类中药黄芪国内外研究的整体态势与热点演变,为其在药食同源领域的深度开发提供理论依据和方向参考。方法基于中国知网(China National Knowledge Infrastructure,CNKI)与科学引文索引(Web of Science,WOS)核心数据...目的系统梳理药食同源类中药黄芪国内外研究的整体态势与热点演变,为其在药食同源领域的深度开发提供理论依据和方向参考。方法基于中国知网(China National Knowledge Infrastructure,CNKI)与科学引文索引(Web of Science,WOS)核心数据库,检索黄芪相关研究文献,最终纳入12881篇中文文献和3378篇英文文献。运用Excel、VOSviewer和CiteSpace等工具,从发文趋势、机构与作者合作、关键词共现、聚类分析及突现分析等多个维度进行文献计量学分析。结果黄芪相关的研究发文量持续增长,中文文献长期占主导地位,侧重于临床应用与功效传承;英文文献自2012年起呈指数级增长,研究重点集中于活性成分及其分子机制。国内外研究机构均以中国中医药高等院校为核心形成集群,中国机构在国际合作中处于枢纽位置,与美国、韩国合作密切。关键词分析显示,研究范式已从单一成分分析转向“多成分-多靶点-多通路”的系统整合研究。值得注意的是,与黄芪药食同源属性直接相关的关键词在当前研究中尚未形成显著独立热点。结论黄芪研究国际化进程加速,已形成以中国为核心、覆盖“基础-临床-产业”全链条的研究体系。但黄芪药食同源应用研究不足,未来应基于理论传承与国际合作,加强其作为食品资源的各关键环节的研究,发掘黄芪在医疗与营养方面的协同价值。展开更多
文摘In daily life,keyword spotting plays an important role in human-computer interaction.However,noise often interferes with the extraction of time-frequency information,and achieving both computational efficiency and recognition accuracy on resource-constrained devices such as mobile terminals remains a major challenge.To address this,we propose a novel time-frequency dual-branch parallel residual network,which integrates a Dual-Branch Broadcast Residual module and a Time-Frequency Coordinate Attention module.The time-domain and frequency-domain branches are designed in parallel to independently extract temporal and spectral features,effectively avoiding the potential information loss caused by serial stacking,while enhancing information flow and multi-scale feature fusion.In terms of training strategy,a curriculum learning approach is introduced to progressively improve model robustness fromeasy to difficult tasks.Experimental results demonstrate that the proposed method consistently outperforms existing lightweight models under various signal-to-noise ratio(SNR)conditions,achieving superior far-field recognition performance on the Google Speech Commands V2 dataset.Notably,the model maintains stable performance even in low-SNR environments such as–10 dB,and generalizes well to unseen SNR conditions during training,validating its robustness to novel noise scenarios.Furthermore,the proposed model exhibits significantly fewer parameters,making it highly suitable for deployment on resource-limited devices.Overall,the model achieves a favorable balance between performance and parameter efficiency,demonstrating strong potential for practical applications.
基金supported by the Natural Science Foundation of Guangxi Province(No.60961002)
文摘In this paper, the specific application of key words Spotting used in the network monitoring is studied, and the keywords spotting is emphasized. The whole monitoring system is divided into two mod-ules： network monitoring and keywords spotting. In the part of network monitoring, this paper adopts a method which is based on ARP spoofing technology to monitor the users＇ data, and to obtain the original audio streams. In the part of keywords spotting, the extraction methods of PLP （one of the main characteristic arameters） is studied, and improved feature parameters- PMCC are put forward. Meanwhile, in order to accurately detect syllable, the paper the double-threshold method with variance of frequency band method, and use the latter to carry out endpoint detection. Finally, keywords recognition module is built by HMM, and identification results are contrasted under Matlab environment. From the experiment results, a better solution for the application of key words recognition technology in network monitoring is found.
文摘This paper focuses on developing a system that allows presentation authors to effectively retrieve presentation slides for reuse from a large volume of existing presentation materials. We assume episodic memories of the authors can be used as contextual keywords in query expressions to efficiently dig out the expected slides for reuse rather than using only the part-of-slide-descriptions-based keyword queries. As a system, a new slide repository is proposed, composed of slide material collections, slide content data and pieces of information from authors＇ episodic memories related to each slide and presentation together with a slide retrieval application enabling authors to use the episodic memories as part of queries. The result of our experiment shows that the episodic memory-used queries can give more discoverability than the keyword-based queries. Additionally, an improvement model is discussed on the slide retrieval for further slide-finding efficiency by expanding the episodic memories model in the repository taking in the links with the author-and-slide-related data and events having been post on the private and social media sites.
基金sponsored by the National Natural Science Foundation of China under grant number No.61100008,61201084the China Postdoctoral Science Foundation under Grant No.2013M541346+3 种基金Heilongiiang Postdoctoral Special Fund(Postdoctoral Youth Talent Program)under Grant No.LBH-TZ0504Heilongjiang Postdoctoral Fund under Grant No.LBH-Z13058the Natural Science Foundation of Heilongjiang Province of China under Grant No.QC2015076The Fundamental Research Funds for the Central Universities of China under grant number HEUCF100602
文摘The e-mail network is a type of social network. This study analyzes user behavior in e-mail subject participation in organizations by using social network analysis. First, the Enron dataset and the position-related information of an employee are introduced, and methods for deletion of false data are presented. Next, the three-layer model(User, Subject, Keyword) is proposed for analysis of user behavior. Then, the proposed keyword selection algorithm based on a greedy approach, and the influence and propagation of an e-mail subject are defined. Finally, the e-mail user behavior is analyzed for the Enron organization. This study has considerable significance in subject recommendation and character recognition.
基金the National Natural Science Foundation of China Grant 71673131 for financial support
文摘Purpose:To reveal the research hotpots and relationship among three research hot topics in b iomedicine,namely CRISPR,iPS(induced Pluripotent Stem)cell and Synthetic biology.Design/methodology/approach:We set up their keyword co-occurrence networks with using three indicators and information visualization for metric analysis.Findings:The results reveal the main research hotspots in the three topics are different,but the overlapping keywords in the three topics indicate that they are mutually integrated and interacted each other.Research limitations:All analyses use keywords,without any other forms.Practical implications:We try to find the information distribution and structure of these three hot topics for revealing their research status and interactions,and for promoting biomedical developments.Originality/value:We chose the core keywords in three research hot topics in biomedicine by using h-index.
文摘[目的]分析中医药治疗多发性硬化(multiple sclerosis,MS)的研究现状及热点,为其临床决策、后续研究及中西医结合诊疗指南构建提供参考。[方法]以中国知网、维普、万方、PubMed、Web of Science数据库为文献来源,检索2004年1月至2024年11月收录的相关文献。采用CiteSpace 6.3.R1软件对文献发表时间、作者、机构、关键词进行共现、聚类及突现等可视化分析,得出其研究现状及热点。[结果]共纳入文献168篇,中医药治疗MS领域的发文量近年无明显变化,而较高质量的临床研究占比增多;目前尚未形成稳定的核心作者群,研究机构集中在各大高校附属医院及中医药院校;高频和高中心性关键词包括辨证论治、名医经验、中西医结合、病因病机和中医证候。[结论]中医药治疗MS处于发展阶段,重点在病因病机及辨证论治研究,未来建议加强团队与机构合作,发展多中心网络,探索多学科交叉、多手段治疗,提高研究质量,为中医药治疗MS的临床决策提供高质量循证证据,以期完善治疗方案,构建中西医结合诊疗指南。
文摘目的系统梳理药食同源类中药黄芪国内外研究的整体态势与热点演变,为其在药食同源领域的深度开发提供理论依据和方向参考。方法基于中国知网(China National Knowledge Infrastructure,CNKI)与科学引文索引(Web of Science,WOS)核心数据库,检索黄芪相关研究文献,最终纳入12881篇中文文献和3378篇英文文献。运用Excel、VOSviewer和CiteSpace等工具,从发文趋势、机构与作者合作、关键词共现、聚类分析及突现分析等多个维度进行文献计量学分析。结果黄芪相关的研究发文量持续增长,中文文献长期占主导地位,侧重于临床应用与功效传承;英文文献自2012年起呈指数级增长,研究重点集中于活性成分及其分子机制。国内外研究机构均以中国中医药高等院校为核心形成集群,中国机构在国际合作中处于枢纽位置,与美国、韩国合作密切。关键词分析显示,研究范式已从单一成分分析转向“多成分-多靶点-多通路”的系统整合研究。值得注意的是,与黄芪药食同源属性直接相关的关键词在当前研究中尚未形成显著独立热点。结论黄芪研究国际化进程加速,已形成以中国为核心、覆盖“基础-临床-产业”全链条的研究体系。但黄芪药食同源应用研究不足,未来应基于理论传承与国际合作,加强其作为食品资源的各关键环节的研究,发掘黄芪在医疗与营养方面的协同价值。
