Current prosthetic control systems explored in the literature that use pattern recognition can perform a limited number of pre-assigned functions, as they must be trained using muscle signals for every movement the us...
详细信息
Current prosthetic control systems explored in the literature that use pattern recognition can perform a limited number of pre-assigned functions, as they must be trained using muscle signals for every movement the user wants to perform. The goal of this study was to explore the development of a prosthetic control system that can classify both trained and novel gestures, for applications in commercial prosthetic arms. The first objective of this study was to evaluate the feasibility of three different algorithms in classifying raw sEMG data for both trained isometric gestures, and for novel isometric gestures that were not included in the training data set. The algorithms used were; a feedforward multi-layer perceptron (FFMLP), a stacked sparse autoencoder (SSAE), and a convolution neural network (CNN). The second objective is to evaluate the algorithms' abilities to classify novel isometric gestures that were not included in the training data set, and to determine the effect of different gesture combinations on the classification accuracy. The third objective was to predict the binary (flexed/extended) digit positions without training the network using kinematic data from the participants hand. A g-tec USB Biosignal Amplifier was used to collect data from eight differential sEMG channels from 10 able-bodied participants. These participants performed 14 gestures including rest, that involved a variety of discrete finger flexion/extension tasks. Forty seconds of data were collected for each gesture at 1200 Hz from eight bipolar sEMG channels. These 14 gestures were then organized into 20 unique gesture combinations, where each combination consisted of a different sub-set of gestures used for training, and another sub-set used as the novel gestures, which were only used to test the algorithms' predictive capabilities. Participants were asked to perform the gestures in such a way where each digit was either fully flexed or fully extended to the best of their abilities.
Tunnel boring machine is extremely sensitive to geological changes, and the accurate prediction of geological conditions ahead of the tunnel face is helpful for safe and efficient tunneling. Since soft methods can use...
详细信息
Tunnel boring machine is extremely sensitive to geological changes, and the accurate prediction of geological conditions ahead of the tunnel face is helpful for safe and efficient tunneling. Since soft methods can use on-site data to predict geological conditions, they are getting more and more attention. However, there is an imbalance between the machine data and geological data, and current soft methods can only utilize limited machine data with geological labels, limiting the performance of the model. To make full use of the massive unlabeled data and limited labeled data, a novel semi-supervised method is proposed to establish the rock mass type prediction model. In the first step, twenty machine parameters are selected as inputs, and the data preprocessing is performed. Thereafter, a geological feature extractor is established based on the stacked sparse autoencoder and unlabeled machine data. Finally, a feature classifier is obtained based on the deep neural network and labeled geological features to realize the prediction of rock mass type. The on-site data collected from Mumbai metro tunnel was utilized to verify the effectiveness of the proposed method. The results indicate that the unsupervised stacked sparse autoencoder is capable of extracting geological features, and the proposed stacked sparse autoencoder and deep neural network-based semi-supervised method outperforms commonly adopted supervised methods. Its classification performance (F-measure) is 13.84%, 10.29%, 8.71%, 5.23% and 5.13% higher than the support vector machine-based, decision tree-based, K-nearest neighbor-based, random forest-based and deep neural network-based methods, respectively. Therefore, the proposed semi-supervised method can predict the rock mass types ahead of the tunnel face more accurately than the current supervised soft methods.
The integration of an edge-preserving filtering technique in the classification of a hyperspectral image (HSI) has been proven effective in enhancing classification performance. This paper proposes an ensemble strateg...
详细信息
The integration of an edge-preserving filtering technique in the classification of a hyperspectral image (HSI) has been proven effective in enhancing classification performance. This paper proposes an ensemble strategy for HSI classification using an edge-preserving filter along with a deep learning model and edge detection. First, an adaptive guided filter is applied to the original HSI to reduce the noise in degraded images and to extract powerful spectral-spatial features. Second, the extracted features are fed as input to a stacked sparse autoencoder to adaptively exploit more invariant and deep feature representations;then, a random forest classifier is applied to fine-tune the entire pretrained network and determine the classification output. Third, a Prewitt compass operator is further performed on the HSI to extract the edges of the first principal component after dimension reduction. Moreover, the regional growth rule is applied to the resulting edge logical image to determine the local region for each unlabeled pixel. Finally, the categories of the corresponding neighborhood samples are determined in the original classification map;then, the major voting mechanism is implemented to generate the final output. Extensive experiments proved that the proposed method achieves competitive performance compared with several traditional approaches. (c) 2017 Society of Photo-Optical Instrumentation Engineers (SPIE)
Deep learning is an important research achievement of artificial intelligence in recent years and has received special attention from scientists around the world. This study applies deep learning to spectral analysis ...
详细信息
Deep learning is an important research achievement of artificial intelligence in recent years and has received special attention from scientists around the world. This study applies deep learning to spectral analysis techniques and proposes a rapid analysis method for cereals. First, the advanced features of the near infrared spectroscopy (NIR) were extracted by the deep learning-stacked sparse autoencoder (SSAE) method, and then the prediction model is built using the affine transformation (AT) and the extreme learning machine (ELM). Experiments were conducted on corn and rice data sets to verify the effectiveness of the method. The results show that the proposed method achieves good prediction results and is superior to other typical NIR analysis methods.
Objectives. Computer-aided pathological voice detection is efficient for initial screening of pathological voice, and has received high academic and clinical attention. This paper proposes an automatic diagnosis metho...
详细信息
Objectives. Computer-aided pathological voice detection is efficient for initial screening of pathological voice, and has received high academic and clinical attention. This paper proposes an automatic diagnosis method of pathological voice based on deep neural network (DNN). Other two classification models (support vector machines and random forests) were used to verify the effectiveness of DNN. Methods. In this paper, we extracted 12 Mel frequency cepstral coefficients of each voice sample as row features. The constructed DNN consists a two-layer stacked sparse autoencoders network and a softmax layer. The stacked sparse autoencoders layer can learn high-level features from raw Mel frequency cepstral coefficients features. Then, the softmax layer can diagnose pathological voice according to high-level features. The DNN and the other two comparison models used the same train set and test set for the experiment. Results. Experimental results reveal that the value of sensitivity, specificity, precision, accuracy, and F1 score of the DNN can reach 97.8%, 99.4%, 99.4%, 98.6%, and 98.4%, respectively. The five indexes of DNN classification results are at least 6.2%, 5%, 5.6%, 5.7%, and 6.2% higher than the comparison models (support vector machine and random forest). Conclusions. The proposed DNN can learn advanced features from raw acoustic features, and distinguish pathological voice from healthy voice. To the extent of this preliminary study, future studies can further explore the application of DNN in other experiments and clinical practice
Milling is a main processing mode of the modern manufacturing industry, which seriously affects the quality and precision of the machined workpiece. However, it is difficult to monitor the tool wear condition in the c...
详细信息
Milling is a main processing mode of the modern manufacturing industry, which seriously affects the quality and precision of the machined workpiece. However, it is difficult to monitor the tool wear condition in the continuous cutting process, especially under a variable speed condition. The existing tool wear condition monitoring methods only carry out analysis with a constant engine speed. Different from the general monitoring methods, this paper put forward a milling cutter wear condition monitoring method based on order analysis (OA) and stacked sparse autoencoder (SSAE). The methodology in the research include signals feature extraction and tool wear state monitoring and were designed to analyze the three-phase spindle current signals instead of the traditional force signals and vibration signals. The variable speed signals were transformed into angle domain stationary signals by order analysis, and the SSAE neural network was used to monitor the tool wear state. The proposed method was verified on the laboratory signals and the results showed a better performance than the other methods and a better applicability in actual industrial manufacturing.
Along with the rapid development of communication network construction, the operation energy consumption grows significantly in recent years, and the expensive electricity cost is hard to be ignored. Therefore, it is ...
详细信息
ISBN:
(纸本)9781509001897
Along with the rapid development of communication network construction, the operation energy consumption grows significantly in recent years, and the expensive electricity cost is hard to be ignored. Therefore, it is necessary to develop an operation energy anomaly detection mechanism to enhance the control ability of electricity cost. According to the practical distribution and data characteristic of smart meters, this paper presents a distributed anomaly detection method of operation energy consumption based on deep learning methods. An IOT-based distributed structure is implemented to execute data interaction. stacked sparse autoencoder is used to extract the high-level representation from massive monitoring data acquired automatically from actual smart meter network. Then softmax is used for classification to detect anomaly and send alarm messages using web technologies. The experimental results show that the proposed method with good prospect for intelligent applications achieves better accuracy and meanwhile decreases computing delay caused by central arithmetic method.
暂无评论