Detection of 14 healthcare-related gestures due to pain at different body parts is the target area of this work using Kinect sensor. the novelty of our work lies in suppressing the problem of compensation by the use o...
详细信息
ISBN:
(纸本)9789811033735;9789811033728
Detection of 14 healthcare-related gestures due to pain at different body parts is the target area of this work using Kinect sensor. the novelty of our work lies in suppressing the problem of compensation by the use of probability while using similarity matching technique for gesture recognition. the adopted method enhances the matching accuracy for all the similarity measures. A shared probability and similarity measure-based metric has been defined as the matching index. this unique technique contributes to field of health care under static gesture recognition as an application of machinelearning with a high accuracy of 99.1071% in 0.0126 s using probability-induced city-block distance.
Intramuscular Electromyography (EMG) signal provides a significant source of information that plays an inevitable role in the diagnosis of neuromuscular disorders. the ensemble method represents a supervised machine l...
详细信息
Efficient and reliable monitoring of wild animals in their natural habitats is essential to inform conservation and management decisions. Automatic covert cameras or "camera traps" are being an increasingly ...
详细信息
ISBN:
(纸本)9781509050048
Efficient and reliable monitoring of wild animals in their natural habitats is essential to inform conservation and management decisions. Automatic covert cameras or "camera traps" are being an increasingly popular tool for wildlife monitoring due to their effectiveness and reliability in collecting data of wildlife unobtrusively, continuously and in large volume. However, processing such a large volume of images and videos captured from camera traps manually is extremely expensive, time-consuming and also monotonous. this presents a major obstacle to scientists and ecologists to monitor wildlife in an open environment. Leveraging on recent advances in deep learning techniques in computer vision, we propose in this paper a framework to build automated animal recognition in the wild, aiming at an automated wildlife monitoring system. In particular, we use a single-labeled dataset from Wildlife Spotter project, done by citizen scientists, and the state-of-the-art deep convolutional neural network architectures, to train a computational system capable of filtering animal images and identifying species automatically. Our experimental results achieved an accuracy at 96.6% for the task of detecting images containing animal, and 90.4% for identifying the three most common species among the set of images of wild animals taken in South-central Victoria, Australia, demonstrating the feasibility of building fully automated wildlife observation. this, in turn, can therefore speed up research findings, construct more efficient citizen science based monitoring systems and subsequent management decisions, having the potential to make significant impacts to the world of ecology and trap camera images analysis.
作者：
Lu, SiyuanWang, HainanWu, XueyanWang, ShuihuaNanjing Normal Univ
Sch Comp Sci & Technol Nanjing 210023 Jiangsu Peoples R China Zhejiang Univ
State Key Lab CAD & CG Hangzhou 310027 Zhejiang Peoples R China Jilin Univ
Minist Educ Key Lab Symbol Computat & Knowledge Engn Changchun 130012 Jilin Peoples R China State Stat Bur
Key Lab Stat Informat Technol & Data Min Chengdu 610225 Sichuan Peoples R China CUNY
City Coll New York Dept Elect Engn New York NY 10031 USA
Magnetic resonance imaging (MRI) is a kind of imaging modality, which offers clearer images of soft tissues than computed tomography (CT). It is especially suitable for brain disease detection. It is beneficial to det...
详细信息
ISBN:
(纸本)9781509034840
Magnetic resonance imaging (MRI) is a kind of imaging modality, which offers clearer images of soft tissues than computed tomography (CT). It is especially suitable for brain disease detection. It is beneficial to detect diseases automatically and accurately. We proposed a pathological brain detection method based on brain MR images and online sequential extreme learningmachine. First, seven wavelet entropies (WE) were extracted from each brain MR image to form the feature vector. then, an online sequential extreme learningmachine (OS-ELM) was trained to differentiate pathological brains from the healthy controls. the experiment results over 132 brain MRIs showed that the proposed approach achieved a sensitivity of 93.51%, a specificity of 92.22%, and an overall accuracy of 93.33%, which suggested that our method is effective.
Proton Magnetic Resonance Spectroscopy (H-1 MRS) has proven its diagnostic potential in a variety of conditions. However, MRS is not yet widely used in clinical routine because of the lack of experts on its diagnostic...
详细信息
ISBN:
(纸本)9783319317441;9783319317434
Proton Magnetic Resonance Spectroscopy (H-1 MRS) has proven its diagnostic potential in a variety of conditions. However, MRS is not yet widely used in clinical routine because of the lack of experts on its diagnostic interpretation. Although data-based decision support systems exist to aid diagnosis, they often take for granted that the data is of good quality, which is not always the case in a real application context. Systems based on models built with bad quality data are likely to underperform in their decision support tasks. In this study, we propose a system to filter out such bad quality data. It is based on convex Non-Negative Matrix Factorization models, used as a dimensionality reduction procedure, and on the use of several classifiers to discriminate between good and bad quality data.
Sparse representation has been successfully used in patternrecognition and machinelearning. However, most existing sparse representation based classification(SRC) methods are to achieve the highest classification ac...
详细信息
ISBN:
(纸本)9781509012572
Sparse representation has been successfully used in patternrecognition and machinelearning. However, most existing sparse representation based classification(SRC) methods are to achieve the highest classification accuracy, assuming the same losses for different misclassifications. this assumption, however, may not hold in many practical applications as different types of misclassification could lead to different losses. To address this problem, we propose a novel cost-sensitive sparse representation based classification(CSSRC) method by using probabilistic modeling. Unlike traditional methods, we predict the class label of test samples by minimizing the misclassification losses, which are obtained via computing the posterior probabilities. Experimental results on the UCI databases validate the efficacy of the proposed approach on average misclassification cost, positive class misclassification rate and negative class misclassification rate. In addition, the experiments show that our proposed method performs competitively compared to SRC, CSSVM and CS4 VM.
Big Data analytics is the core engine in modern decision-making environments. In this talk, we discuss modern signal processing, machinelearning, patternrecognition and optimization tools recently developed and used...
Big Data analytics is the core engine in modern decision-making environments. In this talk, we discuss modern signal processing, machinelearning, patternrecognition and optimization tools recently developed and used in multimedia data bases with a special emphasis on Big multimedia Data. Classification and search in large media repositories will be the main targeted *** talk deals with a new paradigm for multimedia search based on *** present an alternative approach to classical search engines for information retrieval,which can be used for Big and generic multimedia *** introduce an incremental evolution scheme within a collective network of（evolutionary）binary classifier（CNBC）*** proposed framework addresses the problems of feature/class scalability and achieves high classification and content-based retrieval performances over dynamic image *** secret behind the success of CNBC is a novel design to implement the backbone of CNBC,namely the binary *** is a special neural network,which is optimally designed using the recently developed evolutionary optimization algorithm called multi-dimensional particle swarm *** swarm optimization（PSO）is population based stochastic search and optimization process,which was introduced in 1995 by Kennedy and *** goal is to converge to the global optimum of some multi-dimensional fitness *** novel techniques,which extend the basic PSO algorithm,are *** a multidimensional search space where the optimum dimension is unknown,swarm particles can seek both positional and dimensional *** resulting MD-PSO plays a key role in developing machinelearning tools for data classification and feature *** content-based multimedia search engines available today rely heavily on low-level ***,such features extracted automatically usually lack discrimination power needed for accurate description of the image c
Crowd movement analysis is an important issue in social design. this paper studies an machinelearning approach to crowd movement estimation through face image recognition. Although high performance face recognition i...
详细信息
ISBN:
(纸本)9781467369022
Crowd movement analysis is an important issue in social design. this paper studies an machinelearning approach to crowd movement estimation through face image recognition. Although high performance face recognition is a powerful tool in individual authentication with surveillance camera images in public spaces, utilization of personal information is often hesitated under fear of privacy violation. In this paper, a privacy preserving framework for crowd movement analysis is proposed considering k-anonymization of face image features. k-anonymity is a quantitative measure of secureness in data mining and is expected to enhance the utility of personal information. An experimental result demonstrates the applicability of the secure framework in capturing crowd movement characteristics even if individual features are k-aonymized so that each individual is not distinguishable from others k - 1 ones.
Digital circuits are preferred over its analog counterpart withthe invention of microprocessors, microcontrollers, digital signal processors and Field Programmable Gate Arrays (FPGA). Digital circuits in the form of ...
详细信息
ISBN:
(纸本)9781479930708
Digital circuits are preferred over its analog counterpart withthe invention of microprocessors, microcontrollers, digital signal processors and Field Programmable Gate Arrays (FPGA). Digital circuits in the form of digital arithmetic and digital logics are employed for various applications. On the other side, Support Vector machine (SVM) is considered as a state-of-the-art tool for patternrecognition. In this paper, a novel attempt is made to study, assess and design digital circuits using two-class and multi class support vector machine classifiers. It is reported in literature that computational complexity and the classification time for SVM classifier depends on the number of support vectors required for classification. In order to reduce the number of support vectors, an optimum threshold technique based SVM classifier is employed for the design of digital circuits and it is observed that a maximum of 87.5% of total support vectors are pruned on using the technique.
Information fusion is becoming state-of-the-art methodology for performance assessment of engineering assets. Efficiently and smartly combining multi-source information and relevant models from the interested object, ...
详细信息
ISBN:
(纸本)9788895608242
Information fusion is becoming state-of-the-art methodology for performance assessment of engineering assets. Efficiently and smartly combining multi-source information and relevant models from the interested object, more accurate and reliable diagnostic and prognostic results regarding the object can be achieved, which are especially significant for the condition-based maintenance and prognostics and health management applications. Ensemble learning, as a typical machinelearning and decision fusion method, has long been applied in the patternrecognition field and demonstrated promising performance. However, scarce applications of ensemble learning have been found for remaining useful life (RUL) predictions. RUL prediction based on ensemble learning by merging multi-piece information and dynamically updating is proposed in this paper. Specifically, multiple base learners are trained to work as one RUL estimator and weighted averaging with dynamically updated weights upon the latest condition monitoring information is employed to aggregate these RULs to form the final RUL. Rolling element bearing degradation experimental data is used to verify and validate the effectiveness of the proposed method.
暂无评论