Abnormal Event Detection in a Surveillance Scene Using Convolutional Neural Network

Automatic abnormal event detection in a surveillance scene is very significant because of more consciousness about public safety. Because of usefulness and complexity, currently, it is an open research area. In this manuscript, the authors have proposed a novel convolutional neural network (CNN) model to detect an abnormal event in a surveillance scene. In this work, CNN is used in two ways. Firstly, it is used for both feature extraction and classification. In a second way, CNN is used for feature extraction, and support vector machine (SVM) is used for classification. Without any pre-processing, the proposed model gives better results compared to state-of-the-art methods. Experiments are carried out on four different publicly available benchmark datasets and one combined dataset, which contains all images of four datasets. The performance is measured by accuracy and area under the ROC (receiver operating characteristic) curve (AUC). The experimental results determine the efficacy of the proposed model.

Download Full-text

Fault Diagnosis and Fault Frequency Determination of Permanent Magnet Synchronous Motor Based on Deep Learning

Sensors ◽

10.3390/s21113608 ◽

2021 ◽

Vol 21 (11) ◽

pp. 3608

Author(s):

Chiao-Sheng Wang ◽

I-Hsi Kao ◽

Jau-Woei Perng

Keyword(s):

Neural Network ◽

Feature Extraction ◽

Deep Learning ◽

Convolutional Neural Network ◽

Permanent Magnet ◽

Support Vector ◽

One Dimensional ◽

Proposed Model ◽

Wide Range ◽

Fault State

The early diagnosis of a motor is important. Many researchers have used deep learning to diagnose motor applications. This paper proposes a one-dimensional convolutional neural network for the diagnosis of permanent magnet synchronous motors. The one-dimensional convolutional neural network model is weakly supervised and consists of multiple convolutional feature-extraction modules. Through the analysis of the torque and current signals of the motors, the motors can be diagnosed under a wide range of speeds, variable loads, and eccentricity effects. The advantage of the proposed method is that the feature-extraction modules can extract multiscale features from complex conditions. The number of training parameters was reduced so as to solve the overfitting problem. Furthermore, the class feature map was proposed to automatically determine the frequency component that contributes to the classification using the weak learning method. The experimental results reveal that the proposed model can effectively diagnose three different motor states—healthy state, demagnetization fault state, and bearing fault state. In addition, the model can detect eccentric effects. By combining the current and torque features, the classification accuracy of the proposed model is up to 98.85%, which is higher than that of classical machine-learning methods such as the k-nearest neighbor and support vector machine.

Download Full-text

Deep Learning Assisted Neonatal Cry Classification via Support Vector Machine Models

Frontiers in Public Health ◽

10.3389/fpubh.2021.670352 ◽

2021 ◽

Vol 9 ◽

Author(s):

Ashwini K ◽

P. M. Durai Raj Vincent ◽

Kathiravan Srinivasan ◽

Chuan-Yu Chang

Keyword(s):

Neural Network ◽

Machine Learning ◽

Support Vector Machine ◽

Feature Extraction ◽

Deep Learning ◽

Convolutional Neural Network ◽

Support Vector ◽

Svm Classifier ◽

Infant Cry ◽

Learning Techniques

Neonatal infants communicate with us through cries. The infant cry signals have distinct patterns depending on the purpose of the cries. Preprocessing, feature extraction, and feature selection need expert attention and take much effort in audio signals in recent days. In deep learning techniques, it automatically extracts and selects the most important features. For this, it requires an enormous amount of data for effective classification. This work mainly discriminates the neonatal cries into pain, hunger, and sleepiness. The neonatal cry auditory signals are transformed into a spectrogram image by utilizing the short-time Fourier transform (STFT) technique. The deep convolutional neural network (DCNN) technique takes the spectrogram images for input. The features are obtained from the convolutional neural network and are passed to the support vector machine (SVM) classifier. Machine learning technique classifies neonatal cries. This work combines the advantages of machine learning and deep learning techniques to get the best results even with a moderate number of data samples. The experimental result shows that CNN-based feature extraction and SVM classifier provides promising results. While comparing the SVM-based kernel techniques, namely radial basis function (RBF), linear and polynomial, it is found that SVM-RBF provides the highest accuracy of kernel-based infant cry classification system provides 88.89% accuracy.

Download Full-text

Persian Handwritten Digit Recognition Using Combination of Convolutional Neural Network and Support Vector Machine Methods

The International Arab Journal of Information Technology ◽

10.34028/iajit/17/4/16 ◽

2020 ◽

Vol 17 (4) ◽

pp. 572-578

Author(s):

Mohammad Parseh ◽

Mohammad Rahmanimanesh ◽

Parviz Keshavarzi

Keyword(s):

Neural Network ◽

Support Vector Machine ◽

Feature Extraction ◽

Convolutional Neural Network ◽

Recognition Rate ◽

Support Vector ◽

Svm Classifier ◽

Handwritten Digit Recognition ◽

Digit Recognition ◽

Handwritten Digit

Persian handwritten digit recognition is one of the important topics of image processing which significantly considered by researchers due to its many applications. The most important challenges in Persian handwritten digit recognition is the existence of various patterns in Persian digit writing that makes the feature extraction step to be more complicated.Since the handcraft feature extraction methods are complicated processes and their performance level are not stable, most of the recent studies have concentrated on proposing a suitable method for automatic feature extraction. In this paper, an automatic method based on machine learning is proposed for high-level feature extraction from Persian digit images by using Convolutional Neural Network (CNN). After that, a non-linear multi-class Support Vector Machine (SVM) classifier is used for data classification instead of fully connected layer in final layer of CNN. The proposed method has been applied to HODA dataset and obtained 99.56% of recognition rate. Experimental results are comparable with previous state-of-the-art methods

Download Full-text

Detection of Bleeding Events in Electronic Health Record Notes Using Convolutional Neural Network Models Enhanced With Recurrent Neural Network Autoencoders: Deep Learning Approach (Preprint)

10.2196/preprints.10788 ◽

2018 ◽

Author(s):

Rumeng Li ◽

Baotian Hu ◽

Feifan Liu ◽

Weisong Liu ◽

Francesca Cunningham ◽

...

Keyword(s):

Neural Network ◽

Electronic Health Record ◽

Convolutional Neural Network ◽

Language Processing ◽

Characteristic Curve ◽

Bleeding Event ◽

Support Vector ◽

Health Record ◽

Bleeding Events ◽

Electronic Health

BACKGROUND Bleeding events are common and critical and may cause significant morbidity and mortality. High incidences of bleeding events are associated with cardiovascular disease in patients on anticoagulant therapy. Prompt and accurate detection of bleeding events is essential to prevent serious consequences. As bleeding events are often described in clinical notes, automatic detection of bleeding events from electronic health record (EHR) notes may improve drug-safety surveillance and pharmacovigilance. OBJECTIVE We aimed to develop a natural language processing (NLP) system to automatically classify whether an EHR note sentence contains a bleeding event. METHODS We expert annotated 878 EHR notes (76,577 sentences and 562,630 word-tokens) to identify bleeding events at the sentence level. This annotated corpus was used to train and validate our NLP systems. We developed an innovative hybrid convolutional neural network (CNN) and long short-term memory (LSTM) autoencoder (HCLA) model that integrates a CNN architecture with a bidirectional LSTM (BiLSTM) autoencoder model to leverage large unlabeled EHR data. RESULTS HCLA achieved the best area under the receiver operating characteristic curve (0.957) and F1 score (0.938) to identify whether a sentence contains a bleeding event, thereby surpassing the strong baseline support vector machines and other CNN and autoencoder models. CONCLUSIONS By incorporating a supervised CNN model and a pretrained unsupervised BiLSTM autoencoder, the HCLA achieved high performance in detecting bleeding events.

Download Full-text

Super-resolution Ultrasound Imaging Scheme Based on a Symmetric Series Convolutional Neural Network

10.36227/techrxiv.15066315 ◽

2021 ◽

Author(s):

Lakpa Dorje Tamang

Keyword(s):

Neural Network ◽

Feature Extraction ◽

Convolutional Neural Network ◽

Ultrasound Image ◽

Similarity Index ◽

Super Resolution ◽

Input Image ◽

Feature Maps ◽

Proposed Model ◽

Reconstruction Performance

In this paper, we propose a symmetric series convolutional neural network (SS-CNN), which is a novel deep convolutional neural network (DCNN)-based super-resolution (SR) technique for ultrasound medical imaging. The proposed model comprises two parts: a feature extraction network (FEN) and an up-sampling layer. In the FEN, the low-resolution (LR) counterpart of the ultrasound image passes through a symmetric series of two different DCNNs. The low-level feature maps obtained from the subsequent layers of both DCNNs are concatenated in a feed forward manner, aiding in robust feature extraction to ensure high reconstruction quality. Subsequently, the final concatenated features serve as an input map to the latter 2D convolutional layers, where the textural information of the input image is connected via skip connections. The second part of the proposed model is a sub-pixel convolutional (SPC) layer, which up-samples the output of the FEN by multiplying it with a multi-dimensional kernel followed by a periodic shuffling operation to reconstruct a high-quality SR ultrasound image. We validate the performance of the SS-CNN with publicly available ultrasound image datasets. Experimental results show that the proposed model achieves an exquisite reconstruction performance of ultrasound image over the conventional methods in terms of peak signal-to-noise ratio (PSNR), and structural similarity index (SSIM), while providing compelling SR reconstruction time.

Download Full-text

Diagnosis of Alzheimer’s Disease Severity with fMRI Images Using Robust Multitask Feature Extraction Method and Convolutional Neural Network (CNN)

Computational and Mathematical Methods in Medicine ◽

10.1155/2021/5514839 ◽

2021 ◽

Vol 2021 ◽

pp. 1-15

Author(s):

Morteza Amini ◽

MirMohsen Pedram ◽

AliReza Moradi ◽

Mahshad Ouchani

Keyword(s):

Neural Network ◽

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Feature Extraction ◽

Convolutional Neural Network ◽

Nearest Neighbor ◽

Learning Algorithm ◽

Early Stage ◽

Support Vector ◽

Feature Extraction Method

The automatic diagnosis of Alzheimer’s disease plays an important role in human health, especially in its early stage. Because it is a neurodegenerative condition, Alzheimer’s disease seems to have a long incubation period. Therefore, it is essential to analyze Alzheimer’s symptoms at different stages. In this paper, the classification is done with several methods of machine learning consisting of K -nearest neighbor (KNN), support vector machine (SVM), decision tree (DT), linear discrimination analysis (LDA), and random forest (RF). Moreover, novel convolutional neural network (CNN) architecture is presented to diagnose Alzheimer’s severity. The relationship between Alzheimer’s patients’ functional magnetic resonance imaging (fMRI) images and their scores on the MMSE is investigated to achieve the aim. The feature extraction is performed based on the robust multitask feature learning algorithm. The severity is also calculated based on the Mini-Mental State Examination score, including low, mild, moderate, and severe categories. Results show that the accuracy of the KNN, SVM, DT, LDA, RF, and presented CNN method is 77.5%, 85.8%, 91.7%, 79.5%, 85.1%, and 96.7%, respectively. Moreover, for the presented CNN architecture, the sensitivity of low, mild, moderate, and severe status of Alzheimer patients is 98.1%, 95.2%,89.0%, and 87.5%, respectively. Based on the findings, the presented CNN architecture classifier outperforms other methods and can diagnose the severity and stages of Alzheimer’s disease with maximum accuracy.

Download Full-text

Classification of clothing images based on a parallel convolutional neural network and random vector functional link optimized by the grasshopper optimization algorithm

Textile Research Journal ◽

10.1177/00405175211059207 ◽

2021 ◽

pp. 004051752110592

Author(s):

Zhiyu Zhou ◽

Wenxiong Deng ◽

Yaming Wang ◽

Zefei Zhu

Keyword(s):

Neural Network ◽

Feature Extraction ◽

Convolutional Neural Network ◽

Random Vector ◽

Optimization Algorithm ◽

Activation Function ◽

Support Vector ◽

Functional Link ◽

Grasshopper Optimization Algorithm ◽

Grasshopper Optimization

To improve accuracy in clothing image recognition, this paper proposes a clothing classification method based on a parallel convolutional neural network (PCNN) combined with an optimized random vector functional link (RVFL). The method uses the PCNN model to extract features of clothing images. Then, the structure-intensive and dual-channel convolutional neural network (i.e., the PCNN) is used to solve the problems of traditional convolutional neural networks (e.g., limited data and prone to overfitting). Each convolutional layer is followed by a batch normalization layer, and the leaky rectified linear unit activation function and max-pooling layers are used to improve the performance of the feature extraction. Then, dropout layers and fully connected layers are used to reduce the amount of calculation. The last layer uses the RVFL as optimized by the grasshopper optimization algorithm to replace the SoftMax layer and classify the features, further improving the stability and accuracy of classification. In this study, two aspects of the classification (feature extraction and feature classification) are improved, effectively improving the accuracy. The experimental results show that on the Fashion-Mnist dataset, the accuracy of the algorithm in this study reaches 92.93%. This value is 1.36%, 2.05%, 0.65%, and 3.76% higher than that of the local binary pattern (LBP)-support vector machine (SVM), histogram of oriented gradients (HOG)-SVM, LBP-HOG-SVM, and AlexNet-sparse representation-based classifier algorithms, respectively, effectively demonstrating the classification performance of the algorithm.

Download Full-text

Improving Urban Land Cover/Use Mapping by Integrating A Hybrid Convolutional Neural Network and An Automatic Training Sample Expanding Strategy

Remote Sensing ◽

10.3390/rs12142292 ◽

2020 ◽

Vol 12 (14) ◽

pp. 2292

Author(s):

Xin Luo ◽

Xiaohua Tong ◽

Zhongwen Hu ◽

Guofeng Wu

Keyword(s):

Neural Network ◽

Feature Extraction ◽

Land Cover ◽

Convolutional Neural Network ◽

Training Sample ◽

Urban Land ◽

Support Vector ◽

Land Cover Mapping ◽

Urban Regions ◽

Urban Land Cover

Moderate spatial resolution (MSR) satellite images, which hold a trade-off among radiometric, spectral, spatial and temporal characteristics, are extremely popular data for acquiring land cover information. However, the low accuracy of existing classification methods for MSR images is still a fundamental issue restricting their capability in urban land cover mapping. In this study, we proposed a hybrid convolutional neural network (H-ConvNet) for improving urban land cover mapping with MSR Sentinel-2 images. The H-ConvNet was structured with two streams: one lightweight 1D ConvNet for deep spectral feature extraction and one lightweight 2D ConvNet for deep context feature extraction. To obtain a well-trained 2D ConvNet, a training sample expansion strategy was introduced to assist context feature learning. The H-ConvNet was tested in six highly heterogeneous urban regions around the world, and it was compared with support vector machine (SVM), object-based image analysis (OBIA), Markov random field model (MRF) and a newly proposed patch-based ConvNet system. The results showed that the H-ConvNet performed best. We hope that the proposed H-ConvNet would benefit for the land cover mapping with MSR images in highly heterogeneous urban regions.

Download Full-text