检索结果-内蒙古大学图书馆

Optimized hidden markov model based on constrained particle swarm optimization

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Chang, Liu Ouzrout, Yacine Nongaillard, Antoine Bouras, Abdelaziz Jiliu, Zhou DISP Laboratory University Lumiere of Lyon 2 France Key Laboratory of Pattern Recognition and Intelligent Information Processing in Sichuan China College of Information Science and Technology Chengdu University Chengdu China

As one of Bayesian analysis tools, Hidden Markov Model (HMM) has been used to in extensive applications. Most HMMs are solved by Baum-Welch algorithm (BWHMM) to predict the model parameters, which is difficult to find global optimal solutions. This paper proposes an optimized Hidden Markov Model with Particle Swarm Optimization (PSO) algorithm and so is called PSOHMM. In order to overcome the statistical constraints in HMM, the paper develops re-normalization and re-mapping mechanisms to ensure the constraints in HMM. The experiments have shown that PSOHMM can search better solution than BWHMM, and has faster convergence speed. Copyright © 2018, The Authors. All rights reserved.

关键词： Hidden Markov models

Effect of molecular alignment and orientation on elliptically polarized high-order harmonic generation

学校读者我要写书评

暂无评论

Physical Review A 2024年第4期109卷 043118-043118页

作者： Ru Zhang Feng Wang Hua Yuan Shiju Chen Chunhui Yang Tianxin Ou Jiahang Song Qing Liao Peixiang Lu Hubei Key Laboratory of Optical Information and Pattern Recognition Wuhan Institute of Technology Wuhan 430205 China School of Physics and Optoelectronic Engineering Shandong University of Technology Zibo 255049 China Guangdong Intelligent Robotics Institute Dongguan 523808 China

We present a theoretical investigation of the influence of molecular alignment and orientation on elliptically polarized high-order harmonic generation (HHG) from CO molecules exposed to a linearly polarized pulse. The results show that the ellipticity of HHG observed in a partially aligned or oriented CO molecular ensemble significantly deviates from that of the individual response. Additionally, we find that the ellipticity of HHG in the CO molecular ensemble exhibits strong dependence on the degree of molecular alignment/orientation as well as the time delay between the alignment/orientation and probe pulses. More importantly, under the proper alignment/orientation angle, a large HHG ellipticity can be achieved from the partially aligned/oriented molecular ensemble at a moderate degree of alignment/orientation. These findings relax the experimental requirement of a high degree of molecular alignment/orientation for harmonic generation with large ellipticity.

关键词： High-order harmonic generation

Towards multimodal response generation with exemplar augmentation and curriculum optimization

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Lei, Zeyang Li, Zekang Zhang, Jinchao Meng, Fandong Feng, Yang Yang, Yujiu Niu, Cheng Zhou, Jie Tsinghua University China Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Pattern Recognition Center WeChat AI Tencent Inc China

Recently, variational auto-encoder (VAE) based approaches have made impressive progress on improving the diversity of generated responses. However, these methods usually suffer the cost of decreased relevance accompanied by diversity improvements (Zhang et al., 2018b). In this paper, we propose a novel multimodal1 response generation framework with exemplar augmentation and curriculum optimization to enhance relevance and diversity of generated responses. First, unlike existing VAE-based models that usually approximate a simple Gaussian posterior distribution, we present a Gaussian mixture posterior distribution (i.e, multimodal) to further boost response diversity, which helps capture complex semantics of responses. Then, to ensure that relevance does not decrease while diversity increases, we fully exploit similar examples (exemplars) retrieved from the training data into posterior distribution modeling to augment response relevance. Furthermore, to facilitate the convergence of Gaussian mixture prior and posterior distributions, we devise a curriculum optimization strategy to progressively train the model under multiple training criteria from easy to hard. Experimental results on widely used SwitchBoard and DailyDialog datasets demonstrate that our model achieves significant improvements compared to strong baselines in terms of diversity and relevance. Copyright © 2020, The Authors. All rights reserved.

关键词： Curricula

DAmageNet: A universal adversarial dataset

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Chen, Sizhe Huang, Xiaolin He, Zhengbao Sun, Chengjin Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University MOE Key Laboratory of System Control and Information Processing 800 Dongchuan Road Shanghai200240 China

It is now well known that deep neural networks (DNNs) are vulnerable to adversarial attack. Adversarial samples are similar to the clean ones, but are able to cheat the attacked DNN to produce incorrect predictions in high confidence. But most of the existing adversarial attacks have high success rate only when the information of the attacked DNN is well-known or could be estimated by massive queries. A promising way is to generate adversarial samples with high transferability. By this way, we generate 96020 transferable adversarial samples from original ones in ImageNet. The average difference, measured by root means squared deviation, is only around 3.8 on average. However, the adversarial samples are misclassified by various models with an error rate up to 90%. Since the images are generated independently with the attacked DNNs, this is essentially zero-query adversarial attack. We call the dataset DAmageNet, which is the first universal adversarial dataset that beats many models trained in ImageNet. By finding the drawbacks, DAmageNet could serve as a benchmark to study and improve robustness of DNNs. DAmageNet could be downloaded in http://***/Show/56/122 Copyright © 2019, The Authors. All rights reserved.

关键词： Deep neural networks

Detection Level of Apple Based on BP Neural Network

学校读者我要写书评

暂无评论

Detection Level of Apple Based on BP Neural Network

International Conference on Information Sciences,Machinery,Materials and Energy（ICISMME 2015）

作者： Xiaoling LI Jimin YUAN School of Computer Science and Technology Chengdu University University Key Laboratory of Pattern Recognition and Intelligent information Processing The Department Computer Engineering Chengdu Aeronautic Vocational and Technical College

In view of the draw backs of apple grade identification in China,which still relies on photoelectric sorting and manual separation,this paper presents a processing method on the basis of the technology of computer vision and digital *** image processing technology,the researcher calculated the length of the long-short-axis,marked the location of it and calculated the 4 parameters,color,mean square,shape,size,as the key characteristics of the BP input of network to build a network and identify the level of apple through analysis of the external characteristics of *** optimum structure parameters of the BP neural network which had 9 hidden layer neurons were determined by RP training *** showed that average accuracy for fruit classification can reach 92.5% by using this model,and the executing time of microcomputer for grading of one apple is 9.3 *** method has the characteristics of high accuracy and good real-time performance.

关键词： characteristic parameter BP neural network apple detect level

QT-based monitoring system of multi-functional laboratory

学校读者我要写书评

暂无评论

Research Journal of Applied Sciences, Engineering and Technology 2012年第21期4卷 4480-4483页

作者： Xiaoling, Li Jimin, Yuan School of Computer Science and Technology Chengdu University Chengdu 610106 China Key University Key Laboratory of Pattern Recognition and Intelligent information Processing Sichuan Province China School of Computer Science and Technology Panzhihua University Panzhihua 617000 China

Regarding the embedded processor as the core, this study utilizes various cutting-edge technologies such as wireless LAN, USB interface, Bluetooth, multimedia, etc., to propose the design program of QT-based security monitoring system. Taking the lab environment in school as an example, this system has achieved the security monitoring, information transmission and control of certain equipment. Besides, it has cut the Linux kernel module reasonably and explored the touch screen, serial port, wireless LAN, Bluetooth, USB and other resources, thus realizing various functions, such as collection and processing of audio, video and security information and wireless communication. Thereby, users can carry out real-time monitoring for multiple locations through the wireless LAN. © Maxwell Scientific Organization, 2012.

关键词： Bluetooth

A novel feature extracting method for dynamic gesture recognition based on support vector machine

学校读者我要写书评

暂无评论

A novel feature extracting method for dynamic gesture recogn...

International Conference on Information and Automation (ICIA)

作者： Yuanrong Xu Qianqian Wang Xiao Bai Yen-Lun Chen Xinyu Wu Shenzhen Key Lab for Computer Vision and Pattern Recognition Chinese Academy of Sciences University of Science and Technology of China Dept. Mechanical and Automation Engineering The Chinese University of Hong Kong Guangdong Provincial Key Laboratory of Robotics and Intelligent System Chinese Academy of Sciences

In this paper, we propose a method based on the SVM algorithm to recognize dynamic hand gestures. The information of motion trajectory is captured by a leap motion in three-dimension space. A new methodology of feature extracting is proposed to guarantee the length of samples being the same. The elements of feature vectors are ranged according to two different criteria: one is the amplitude of the variation of orientation angles, and the other criterion is the order of the appearance of features. Experimental results show that this method can classify the dynamic hand gestures effectively.

关键词： Vectors Trajectory Feature extraction Equations Mathematical model Hidden Markov models Automation

A Multi Modal Content-Based Copy Detection Approach

学校读者我要写书评

暂无评论

A Multi Modal Content-Based Copy Detection Approach

International Conference on Computational Intelligence and Security

作者： Siyuan Wu Zhicheng Zhao Multimedia Communication and Pattern Recognition Labs Beijing University of Posts and Telecommunications Beijing Key Laboratory of Network System and Network Culture Beijing University of Posts and Telecommunications Beijing China

As an alternative to the watermarking approach, content-based copy detection (CBCD) recently has become a promising technique for video monitoring and copyright protection. In this paper, a novel CBCD framework is proposed. Firstly, robust global feature (DCT), local features (SIFT) and audio feature (EDF) are first combined to describe video and audio contents, and the density sampling method is used to improve the generation of visual codebook. Secondly, picture-in-picture (PIP) detection algorithm is introduced to find the PIP transformation of videos. Meanwhile, a video matching method based on visual codebook is presented to calculate the similarity of copy videos and locality-sensitive hashing local (LSH) method is used for DCT indexing. Finally, a hierarchical fusion scheme is used to refine the detection results. Experiments on TRECVID dataset show that the proposed framework gives better results than the average results of CBCD task in TRECVID 2011.

关键词： Feature extraction Discrete cosine transforms Visualization Indexes Robustness Detection algorithms Fingerprint recognition

Unifying Gradients to Improve Real-World Robustness for Deep Networks

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Wu, Yingwen Chen, Sizhe Fang, Kun Huang, Xiaolin Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University 800 Dongchuan Road Shanghai China Institute of Image Processing and Pattern Recognition The MOE Key Laboratory of System Control and Information Processing Shanghai Jiao Tong University 800 Dongchuan Road Shanghai China

The wide application of deep neural networks (DNNs) demands an increasing amount of attention to their real-world robustness, i.e., whether a DNN resists black-box adversarial attacks, among which score-based query attacks (SQAs) are most threatening since they can effectively hurt a victim network with the only access to model outputs. Defending against SQAs requires a slight but artful variation of outputs due to the service purpose for users, who share the same output information with SQAs. In this paper, we propose a real-world defense by Unifying Gradients (UniG) of different data so that SQAs could only probe a much weaker attack direction that is similar for different samples. Since such universal attack perturbations have been validated as less aggressive than the input-specific perturbations, UniG protects real-world DNNs by indicating attackers a twisted and less informative attack direction. We implement UniG efficiently by a Hadamard product module which is plug-and-play. According to extensive experiments on 5 SQAs, 2 adaptive attacks and 7 defense baselines, UniG significantly improves real-world robustness without hurting clean accuracy on CIFAR10 and ImageNet. For instance, UniG maintains a model of 77.80% accuracy under 2500-query Square attack while the state-of-the-art adversarially-trained model only has 67.34% on CIFAR10. Simultaneously, UniG outperforms all compared baselines in terms of clean accuracy and achieves the smallest modification of the model output. The code is released at https://***/snowien/UniG-pytorch. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep neural networks