In our previous work, we developed an IMU (Inertial Measurement Unit) based smart ring that allows users to type characters without a physical keyboard and adopt well-known pattern recognition algorithms, such as Supp...
详细信息
In our previous work, we developed an IMU (Inertial Measurement Unit) based smart ring that allows users to type characters without a physical keyboard and adopt well-known pattern recognition algorithms, such as Support Vector Machine (SVM), and Naive Bayes (NB), for keystroke recognition. However, these algorithms always require intensive computing resources or offer limited recognition accuracy. Moreover, they are often seen as black boxes incapable of providing readily comprehensible and visible clues for classification. This hampers the improvement of keystroke recognition accuracy and the ring-type virtual keyboard's character layout design. Here we present a novel algorithm to recognize keystrokes in a fast and accurate manner. Firstly, the standard feature vector, including five attitude angle features and one acceleration feature, is built to express a specific stroke. Then, the feature vector of the testing keystroke is compared with the standard features. The most similar keystroke is matched and recognized after three times of voting. Based on this algorithm, we can identify the easily confused keystrokes and understand the mechanisms behind it. With this interpretability, we will be able to achieve the customized ring-type virtual keyboard application if necessary. The performance of this algorithm was evaluated by using a dataset with 1500 keystrokes of three different subjects. The results show that our algorithm is more effective in keystroke recognition than traditional algorithms for this ring-type keyboard. In addition to its application on virtual keyboards, this algorithm can also be potentially applied on other classification tasks with easy-to-understand results.
作者：
Takaoka, AsahiKanagawa Univ
Dept Informat Syst Creat Kanagawa Ku Rokkakubashi 3-27-1 Yokohama Kanagawa 2218686 Japan
A simple-triangle graph is the intersection graph of triangles that are defined by a point on a horizontal line and an interval on another horizontal line. The time complexity of the recognition problem for simple-tri...
详细信息
A simple-triangle graph is the intersection graph of triangles that are defined by a point on a horizontal line and an interval on another horizontal line. The time complexity of the recognition problem for simple-triangle graphs was a longstanding open problem, which was recently settled. This paper provides a new recognition algorithm for simple-triangle graphs to improve the time bound from O(n(2)(m) over bar) to O(nm), where n, (m) over bar, and m are the number of vertices, edges, and non-edges of the graph, respectively. The algorithm uses the vertex ordering characterization that a graph is a simple-triangle graph if and only if there is a linear ordering of the vertices containing both an alternating orientation of the graph and a transitive orientation of the complement of the graph. We also show, as a byproduct, that an alternating orientation can be obtained in O(nm) time for cocomparability graphs, and it is NP-complete to decide whether a graph has an orientation that is alternating and acyclic. (C) 2019 Elsevier B.V. All rights reserved.
Because of the hierarchical significance of traffic sign images, the traditional methods do not effectively control and extract the brightness and features of layered images. Therefore, an automatic recognition algori...
详细信息
Because of the hierarchical significance of traffic sign images, the traditional methods do not effectively control and extract the brightness and features of layered images. Therefore, an automatic recognition algorithm for traffic signs based on a convolution neural network is proposed in this paper. First, the histogram equalization method is used to pre-process the traffic sign images, with details of the images being enhanced and contrast of the images improved. Then, the traffic sign images are recognized by a convolution neural network and the large scale structure of information in the traffic sign images are obtained by using a hierarchical significance detection method based on graphical models. Next, the area of interest in the traffic sign images are extracted by using the hierarchical significance model. Finally, the Softmax classifier is selected to classify the input feature images to realize the automatic recognition of traffic signs. Experimental results show that the proposed algorithm can control the brightness of traffic sign images, which can accurately extract image regions of interest and complete the automatic recognition of traffic signs.
The existing studies suggest the following methods for measuring sleep postures: installing cameras and record the sleep postures then analyze the postures or measuring the change of the posture by attaching the senso...
详细信息
ISBN:
(纸本)9781450362870
The existing studies suggest the following methods for measuring sleep postures: installing cameras and record the sleep postures then analyze the postures or measuring the change of the posture by attaching the sensor to the user's body. However, the installation of high-cost devices such as a camera or direct attachment of sensors on the subject's body present issues relating to the cost and convenience. As a solution, this paper develops a recognition algorithm for analyzing sleep postures using a smart pad with embedded fabric type pressure sensors. This algorithm is applied using a fabric-made smart pad with multiple pressure sensors. The sensors detect the distribution of the body pressure on the pad during sleeping, and the collected body pressure distribution data determines the sleep postures of the users. Further, this smart fabric pad does not require any additional analytical devices. This pad allows the users to monitor own sleep postures continuously. With the analyzed sleep posture data, the user can recognize one's typical sleep postures. In order to verify the effectiveness of the algorithm, this paper conducts an experiment to validate using the sleep posture data defined as nine categories. As a result, the algorithm had an average of 91.4% accuracy rate.
Early defibrillation plays very important role to survive patients suffering from sudden cardiac arrest. Thus, there are lots of recognition algorithms for electrocardiogram, and we have also proposed recognition algo...
详细信息
ISBN:
(纸本)9781728121949
Early defibrillation plays very important role to survive patients suffering from sudden cardiac arrest. Thus, there are lots of recognition algorithms for electrocardiogram, and we have also proposed recognition algorithms based on spectrum feature parameters of electrocardiograms (ECGs). In this paper, we show the detailed analysis result of our recognition algorithm, and in particular the degree of similarity between Ventricular Tachycardia (VF) and PEA (Pulseless Electrical Activity) is evaluated.
Power wireless private network has the advantages of high spectrum efficiency, large data throughput and good system stability. However, in practical applications, it may cause large interference to 230 digital transm...
详细信息
ISBN:
(纸本)9781450372985
Power wireless private network has the advantages of high spectrum efficiency, large data throughput and good system stability. However, in practical applications, it may cause large interference to 230 digital transmission stations, which affects the normal operation of the system. Spectrum-sensing technology can be used to capture 230 digital radio signals in the spectrum,so that both systems can work steadily on the same spectrum resources. When it is found that both use the same frequency point at the same time, it is avoided in time to ensure the stable operation of the system. The current communication signal modulation types tend to be diversified and complicated and the signals are more and more dense. Therefore, the perception function of the user is not only to realize the perception of the spectrum signal but also to accurately identify the type of the modulation signal. In view of this, after the cognitive system perceives the presence of authorized user signals in the licensed spectrum, the algorithm proposes a modulation signal recognition technology based on BP neural network, which helps the cognitive radio network of the power wireless private network to better perceive the spectrum environment. Identifying the type of modulation in the spectrum to determine the type of service for the authorized user, thereby facilitating better dynamic use of the spectrum.
Based on four-band Infrared flame detector of 4.26 mu m, 2.2 mu m, 3.9 mu m and 4.8 mu m, the flame characteristic information is extracted and analyzed, and a scheme for realizing the specific identification algorith...
详细信息
ISBN:
(纸本)9781538612446
Based on four-band Infrared flame detector of 4.26 mu m, 2.2 mu m, 3.9 mu m and 4.8 mu m, the flame characteristic information is extracted and analyzed, and a scheme for realizing the specific identification algorithm of four-hand infrared flame detector is proposed. By analyzing and comparing the mathematical relations of the spectral characteristics of the four bands, the identification of the four-band infrared flame is realized by the combination of the threshold method, the mathematical correlation analysis method and the signal average power method. The experimental results show that the recognition algorithm is feasible and reliable, and the accuracy of the infrared flame detector and the ability of adapting to the environment can be improved effectively, and the aim of the high reliability and long-distance detection is realized.
This paper proposes a new recognition algorithm for shockable arrhythmias for patients suffering from sudden cardiac arrest. In this paper, by using gabor wavelet transform (GWT), useful and effective spectrum feature...
详细信息
ISBN:
(纸本)9781509066841
This paper proposes a new recognition algorithm for shockable arrhythmias for patients suffering from sudden cardiac arrest. In this paper, by using gabor wavelet transform (GWT), useful and effective spectrum features for electrocardiogram (ECG) are extracted. The proposed recognition algorithm based on spectrum features can achieve good performance comparing with the existing results, i.e. the result of this paper is useful.
The current target contour recognition algorithm is prone to the problem of insufficient target contour extraction accuracy when the image target edge features are not obvious and the background is complicated. To mak...
详细信息
The current target contour recognition algorithm is prone to the problem of insufficient target contour extraction accuracy when the image target edge features are not obvious and the background is complicated. To make up for the shortcomings of the algorithm, this study proposes an edge-based approach. Sharpened variable operator image target contour recognition algorithm, by introducing Laplacian differential operator and Fourier transform, constructs the edge sharpening operator of the joint image, and then constructs the gradient feature based on the gradient feature. The target contour recognition algorithm of the variable operator image is used to measure the relevant parameters of the manufactured object. The experimental results show that compared with the generalized target contour recognition algorithm, the proposed algorithm can guarantee accuracy and stability in more complex uncertain environments.
The affective computing of music is an essential component of the music artificial intelligence field, especially those regards human-computer interaction technology, as for music mode is a crucial component of musica...
详细信息
ISBN:
(纸本)9789881563958
The affective computing of music is an essential component of the music artificial intelligence field, especially those regards human-computer interaction technology, as for music mode is a crucial component of musical emotion perceptions. To be more specific, the recognition skill of scales, the music mode material remains manually used today, which endow this research a great practical significance for using the algorithm to recognizing them in high speed. This paper describes how to classify the Chinese traditional scales (CTS) and annotate the emotional colors of the Chinese traditional music (CTM) through algorithmrecognition methods, based on setting up a decision tree. This algorithm-testing database consists of MIDI samples either in CTS or not, while taking three steps to determine it: sample pre-processing, decision tree classification, result verification. The experiment results show the accuracy of this cognition method on samples were non restrict to CTS. Which prove the feasibility of indirectly obtaining emotion colors through the process to identify and classify CTM based on the decision tree.
暂无评论