检索结果-内蒙古大学图书馆

International Journal of Automation and computing 2021年第3期18卷 351-376页

作者： Hao Zhu Man-Di Luo Rui Wang Ai-Hua Zheng Ran He Anhui Provincial Key Laboratory of Multimodal Cognitive Computation School of Computer Science and TechnologyAnhui UniversityHefei 230601China Center for Research on Intelligent Perception and Computing(CRIPAC)and National Laboratory of Pattern Recognition(NLPR) Institute of AutomationChinese Academy of SciencesBeijing 100190China School of Artificial Intelligence University of the Chinese Academy of SciencesBeijing 100049China Center for Excellence in Brain Science and Intelligence Technology Chinese Academy of SciencesShanghai 200031China

Audio-visual learning,aimed at exploiting the relationship between audio and visual modalities,has drawn considerable attention since deep learning started to be used *** tend to leverage these two modalities to improve the performance of previously considered single-modality tasks or address new challenging *** this paper,we provide a comprehensive survey of recent audio-visual learning *** divide the current audio-visual learning tasks into four different subfields:audiovisual separation and localization,audio-visual correspondence learning,audio-visual generation,and audio-visual representation ***-of-the-art methods,as well as the remaining challenges of each subfield,are further ***,we summarize the commonly used datasets and challenges.

关键词： Deep audio-visual learning audio-visual separation and localization correspondence learning generative models representation learning

来源：评论

学校读者我要写书评

暂无评论

An improved support vector machine active learning for audio event detection

International Journal of Digital Content Technology and its ...

引用

International Journal of Digital Content Technology and its Applications 2011年第9期5卷 9-16页

作者： Leng, Yan Liu, Gang Guo, Jun Wang, Rongyan Pattern Recognition and Intelligent System Laboratory Beijing University of Posts and Telecommunications Beijing 100876 China College of Physics and Electronics Shandong Normal University Jinan 250014 China

Nowadays, audio event detection is a hot research, the detection needs large amounts of labeled samples for training audio event models, the shortage of such labeled samples is a big obstacle. Currently, active learning is a common method in dealing with the problem of insufficient labeled samples. The most popular support vector machine active learning is to query the sample closest to the current hyperplane, but when the current hyperplane is far away from the true hyperplane, the sample closest to the current hyperplane is not so informative to be a good choice. In order to select more informative samples whenever the current hyperplane is close to or far away from the true hyperplane, this paper proposes margin and misclassification character based sampling (MMCBS) active learning, MMCBS queries the sample that is close to the current hyperplane and meanwhile is more likely to be misclassified. SVM is only interested in support vectors on the boundary, while boundary samples are easy to be misclassified, since MMCBS considers the misclassification character, so the selected samples of MMCBS are more likely to be the informative support vectors. Experiments show that MMCBS achieves higher classification accuracy with less human annotation on the melodrama Friends corpus, it is suitable for classification with scarce labeled data.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Research on DNA sequence homology based on second order Markov model

引用

Journal of Fiber Bioengineering and Informatics 2015年第3期8卷 539-546页

作者： Zhang, Junyan Yang, Chenhui College of Computer Science Chengdu University Chengdu China Key Laboratory of Pattern Recognition and Intelligent Information Processing of Sichuan Chengdu University Chengdu China

DNA sequence homology is a critical and fundamental problem in bioinformatics. In this paper, we solve this problem by use of the second order Markov modal instead of traditional sequence alignment because DNA character sequence meets the Markov properties. Hence, the characteristics of DNA sequences are represented by using their two-step transition probabilities matrices. The similarity degree measurement between two different DNA sequences is defined. Our DSHM algorithm is put forward which is implemented by MyEclipse. The contrast experiments are done between DSHM and other two methods. The experimental results show that DSHM algorithm can determine DNA sequence homology correctly in the more effective way. © 2015 Binary Information Press & Textile Bioengineering and Informatics Society.

关键词： DNA sequences

来源：评论

学校读者我要写书评

暂无评论

Face Forgery Detection Algorithm Based on Improved MobileViT Network 8

Face Forgery Detection Algorithm Based on Improved MobileViT...

引用

8th International Conference on intelligent Computing and Signal Processing, ICSP 2023

作者： Wang, Tiantian Lu, Xiaoqi Inner Mongolia University of Science and Technology School of Information Engineering Baotou014010 China Inner Mongolia Key Laboratory of Pattern Recognition and Intelligent Image Processing Baotou014010 China Inner Mongolia University of Technology Institute of Information Engineering Hohhot010051 China

ISBN: (纸本)9798350302455

DeepFakes blur the boundaries between reality and forgery, resulting in the collapse of exiting credit system, causing immeasurable consequences for national security and social order. Through analysis of existing face forgery techniques, it is found that most generation techniques rely on random noise distribution, and global information will be lost after up sampling. Therefore, we propose a deepfake detection algorithm based on improved MobileViT, which uses CNN local space biasing and the global space representation of the Transformer network to learn the local features and global representation of forged faces, respectively. Coordinate attention is introduced to obtain directional perception and position sensitive information, making the model locate synthetic traces of fake faces better and fusion local and global representation more effectively. For the improved generalization of the model, with the GELU activation function to solve the problem of neuron death. Our model achieved 96.2% on FF++(C23) datasets, and 93.7%,94.1%,96.3%,87.9% on DF, F2F, FS, and NT datasets, respectively. Comparing with previous methods, our model has shown detection robustness and better generalization. © 2023 IEEE.

关键词： National security

来源：评论

学校读者我要写书评

暂无评论

Natural image matting based on image inpainting

Natural image matting based on image inpainting

引用

Computer Graphics, Image and Virtualization (ICCGIV), International Conference on

作者： Yuan Zhang Mian Tan Zhulian Zhou Yuan Yang Yihui Liang Fujian Feng Guizhou Key Laboratory of Pattern Recognition and Intelligent System Guizhou Minzu University Guiyang China School of Computer Science University of Electronic Science and Technology China Zhongshan China

Deep image matting is a hot problem with applications in computer vision and image processing. It has been widely used in image composition, film production and video editing etc. The current matting method based on image inpainting uses a deep neural network to inpaint the foreground target and background region to further improve the accuracy of alpha matte. However, when the trimap contains large unknown regions, the excessive inpainting of foreground and background produce a lot of redundant information, which leads to a degradation of alpha matte quality. Therefore, to address this issue, a natural image matting method based on image inpainting is designed. This method involves the refinement process for trimap, which improves the quality of the trimap, enlarges the foreground and background regions providing additional information for image matting. Extensive experimental results on the composition-1k dataset demonstrate that the presented method provide high-quality alpha mattes not only in the case that the trimap contains small unknown regions, but also in the case that the trimap contains large unknown regions.

关键词： Degradation Deep learning Computer vision Image processing Refining Neural networks Production

来源：评论

学校读者我要写书评

暂无评论

An efficient audio event classification technology for complex audio documents

引用

Journal of Convergence Information Technology 2011年第10期6卷 227-234页

作者： Leng, Yan Liu, Gang Guo, Jun Pattern Recognition and Intelligent System Laboratory Beijing University of Posts and Telecommunications Beijing 100876 China College of Physics and Electronics Shandong Normal University Jinan 250014 China

With the rapid increment of multimedia documents, efficient multimedia document management has become an urgent need. Audio document analysis, as an important way to deal with multimedia document management, has attracted more and more attentions. Audio event classification is a key step of audio document analysis, but most audio event classification technologies are to classify isolated audio events, there are less researches on the classification of overlapped audio events, but in real life, an audio document usually has large parts of overlaps, when dealing with such complex audio documents, the usual audio event classification technologies would have a poor performance. In this paper, taking the film audio documents as the study object, we investigate how to do efficient audio event classification for the complex audio documents, and proposed Pure Model and Clustering Information Classification system (PMCICS). PMCICS can not only improve the classification of overlapped audio events, but also is helpful for the classification of isolated ones, it can improve the overall classification performance, and then can be taken as an efficient scheme for the classification problem of complex audio documents.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Background suppression based-on wavelet transformation to detect infrared target

Background suppression based-on wavelet transformation to de...

引用

International Conference on Machine Learning and Cybernetics, ICMLC 2005

作者： Sun, Y.U.-Qiu Tian, Jin-Wen Liu, Jian School of Information and Mathematics Yangtze University Jingzhou 434023 China State Education Commission Key Laboratory for Image Processing and Intelligent Control Institute for Pattern Recognition and Artificial Intelligence Huazhong University of Science and Technology Wuhan 430074 China State Education Commission Key Laboratory for Image Processing and Intelligent Control Dept. of Electronic Information and Enginery Huazhong University of Science and Technology Wuhan China

ISBN: (纸本)078039092X

Target detection and location in infrared clutter background is very important to infrared search and track system. Especially for small target detection in infrared image in background of sea and sky, there are no geometric and structure character to use. When targets such as ship and naval vessel sailing at long distance, they always appear around the sea-level line, and it is mixed with cloud and sea-wave clutter. It is difficult to segment and locate precisely. Background suppression based-on wavelet transformation is proposed in the paper. Wavelet decomposition makes it possible to analyze a signal both in time and frequency domains. In the paper infrared sea and key background images are processed. For their SNR is low and background is complex, using wavelet transformation decomposes an original image and extract approximate feature to reconstruct an image, which mainly includes background information. A new image will be obtained using background image subtracted from original image. There are mainly target and noise points left in the new image. By setting proper threshold, the target can be detected perfectly. At last, experiment results are given and show the method is practical. © 2005 IEEE.

关键词： pattern recognition

来源：评论

学校读者我要写书评

暂无评论

Research on the digital system of archaeological

Research on the digital system of archaeological

引用

第四届国际遥感考古会议

作者： CAO Ligang WANG Xuben MIAO Fang YI Fasheng Key Laboratory of Earth Exploration & Information Techniques of Ministry of Education Director of Key Lab of Pattern Recognition & Intelligent Information Processing of Sichuan Province

The digital system of Archaeological includes multi-scale non-destructive detection of archaeological methods,data mining technologies and the GIS of archaeological *** preservation is not just for the protection of cultural relics has been *** non-destructive detection method to detect archaeological artifacts and clarify the situation of cultural relics buried *** detection data through the using of data mining algorithm make out the archaeological ***,use GIS technology to achieve the detection of data management and integration of data mining methods.A whole system of digital archaeological built on the GIS platform,based on the using of data mining technology to achieve a detection method and archaeological information mapping,the system for the digital archaeology has provided a complete technical support.

关键词： Archaeology explore GIS Data mining

来源：评论

学校读者我要写书评

暂无评论

A Method for Authenticity Identification of Fritillaria Cirrhosa D. Don Based on Deep Learning 4

A Method for Authenticity Identification of Fritillaria Cirr...

引用

4th IEEE International Conference on Image, Vision and Computing, ICIVC 2019

作者： Hu, Ke Hu, Pan Cao, Dong Yan, Xin Yu, Xi Liu, Chang Chengdu University College of Information Science and Engineering Chengdu China Key Laboratory of Pattern Recognition and Intelligent Information Processing of Sichuan Chengdu University Chengdu China Research Institute of Big Data Chengdu University China Chengdu Institute of Chinese Herbal Medicine China

ISBN: (纸本)9781728123257

Since the authentic Fritillaria Cirrhosa D. Don resources are scarce due to its high price and valuable medical uses, it is difficult to meet the clinical needs. Therefore, the problem of adulteration in the market is more and more serious. At present, the identification of Fritillaria mainly relies on traditional trait identification, microscopic identification, physical and chemical identification and other methods, which is subjective and requires high practical experience for operators, and the pretreatment work is cumbersome. In this paper, the concept of deep learning is introduced into the identification of Fritillaria for the first time. First, we have created the first multi-angle Fritillaria standard datasets, including Fritillaria Cirrhosa D. Don and Fritillary bulb. We then have proposed a novel deep learning framework SE-DPU to classify Fritillaria datasets, from which we can get the highest classification accuracy. The framework is a hybrid of Squeeze-and-Excitation block (SE unit), U-Net and Dual Path Network (DPN). It reuses the features and explores the new features adaptively recalibrates channel-wise feature responses obtains abundance features while training. Experimental results with Fritillaria Cirrhosa D. Don data have indicated that the proposed method have provided competitive performance. © 2019 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Interactive object segmentation using iterative adjustable graph cut

Interactive object segmentation using iterative adjustable g...

引用

2011 IEEE Visual Communications and Image Processing, VCIP 2011

作者： Shi, Ran Liu, Zhi Xue, Yinzhu Zhang, Xiang School of Communication and Information Engineering Shanghai University Shanghai 200072 China Key Laboratory of Advanced Display and System Application Shanghai University Ministry of Education Shanghai 200072 China Institute of Image Processing and Pattern Recognition Shanghai Jiaotong University Shanghai 200240 China

ISBN: (纸本)9781457713200

Interactive object segmentation is widely used for extracting any user-interested objects from natural images. A common problem with many interactive segmentation approaches is that the object segmentation quality is degraded due to inaccurate object/background seeds provided by the user. This paper proposes an iterative adjustable graph cut to efficiently solve this problem. First, object/background seeds are initialized based on the object segmentation result obtained with the user-specified scribbles as the interactive input. Then, an iterative seed adjustment scheme is exploited to correct inaccurate seeds and extract new suitable seeds via graph cut, in which the balancing weight between energy terms are adaptively updated to protect stable seeds and speedup the iteration process. Finally, suitable seeds are obtained and graph cut is used to segment the objects. Experimental results demonstrate the better segmentation performance of our approach even if user provides rather rough seeds. © 2011 IEEE.

关键词： Image segmentation

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：