Research on a Convolution Kernel Initialization Method for Speeding Up the Convergence of CNN

This paper presents a convolution kernel initialization method based on the local binary patterns (LBP) algorithm and sparse autoencoder. This method can be applied to the initialization of the convolution kernel in the convolutional neural network (CNN). The main function of the convolution kernel is to extract the local pattern of the image by template matching as the target feature of subsequent image recognition. In general, the Xavier initialization method and the He initialization method are used to initialize the convolution kernel. In this paper, firstly, some typical sample images were selected from the training set, and the LBP algorithm was applied to extract the texture information of the typical sample images. Then, the texture information was divided into several small blocks, and these blocks were input into the sparse autoencoder (SAE) for pre-training. After finishing the training, the weight values of the sparse autoencoder that met the statistical features of the data set were used as the initial value of the convolution kernel in the CNN. The experimental result indicates that the method proposed in this paper can speed up the convergence of the network in the network training process and improve the recognition rate of the network to an extent.

Multi scale switchable atrous convolution for target detection based on feature pyramid

Repeated observation mechanism can effectively solve the problem of low efficiency of feature extraction. By extracting features for many times to strengthen target features, this paper proposed a multi-scale switchable atrous convolution based on feature pyramid, SPC. The head of the detector adopted pyramid convolution mode, constructs 3-D convolution in the feature pyramid, and detected the same target in different pyramid levels by using the shared convolution with different stride changes, which realized the repeated observation of target features on multi-scale. The module optimized the convolution layer, extracted the features of the same image by convolution check of different sizes, and then selected and integrated the extracted results by using switch function, which effectively expanded the field of view of convolution kernel. In this paper, we choosed retinanet as the baseline network, and improved the loss function of focal loss proposed by retinanet to further solved the problem of unbalanced number of samples and sample distribution in the network model. The proposed method performed well on MS coco data set, improved the average accuracy of 9.8% on the basis of retinanet to 48.9%, and achieved FPS of 5.1 in 1333 * 800 images.

Image extraction using convolution kernal

This paper discusses about the noise reduction of images using the convolution matrix. The convolution kernel matrix filters generate new features of the input images with good quality.The noise reduction methods based on convolution kernel is achieved by deep learning theory along with the difference equations. The random variation of the colour and brightness are taken as authenticated coefficients of the images. Convolution techniques along withrecurrent neuralnetwork are applied into theinput image. This input image is divided into the matrix of pixel values. The optimal enhanced image is arrived through convolution kernel using computational learning of autonomous difference equations.

Expression Recognition Method Using Improved VGG16 Network Model in Robot Interaction

Aiming at the problems of poor representation ability and less feature data when traditional expression recognition methods are applied to intelligent applications, an expression recognition method based on improved VGG16 network is proposed. Firstly, the VGG16 network is improved by using large convolution kernel instead of small convolution kernel and reducing some fully connected layers to reduce the complexity and parameters of the model. Then, the high-dimensional abstract feature data output by the improved VGG16 is input into the convolution neural network (CNN) for training, so as to output the expression types with high accuracy. Finally, the expression recognition method combined with the improved VGG16 and CNN model is applied to the human-computer interaction of the NAO robot. The robot makes different interactive actions according to different expressions. The experimental results based on CK + dataset show that the improved VGG16 network has strong supervised learning ability. It can extract features well for different expression types, and its overall recognition accuracy is close to 90%. Through multiple tests, the interactive results show that the robot can stably recognize emotions and make corresponding action interactions.

Intelligent user identity authentication in vehicle security system based on wireless signals

Intelligent identity authentication in vehicle security systems, as a vital component in anti-theft system and safety driving assist system, has received wide attention. Current vehicle security systems, however, focus the car security on the car keys security, ignore the owner of car keys. Anyone who owns the car keys can operate the car. This paper introduces an intelligent identity authentication method for vehicle security system based on wireless signals. Unlike past work, our approach combines car security with car owners and car keys. The intuition underlying our design is that when a user walks towards the car, the user’s gait information can be leveraged to identify the user. We capture the user’s gait information using wireless devices which can be deployed in the car, and then extract features from the received wireless signals using convolution kernel and apply artificial neural network to identify the user. We built a prototype and experimental results show that our approach can achieve high accuracy and strong robustness.

Deep-Learning-Based Remaining Useful Life Prediction Based on a Multi-Scale Dilated Convolution Network

Remaining useful life (RUL) prediction of key components is an important influencing factor in making accurate maintenance decisions for mechanical systems. With the rapid development of deep learning (DL) techniques, the research on RUL prediction based on the data-driven model is increasingly widespread. Compared with the conventional convolution neural networks (CNNs), the multi-scale CNNs can extract different-scale feature information, which exhibits a better performance in the RUL prediction. However, the existing multi-scale CNNs employ multiple convolution kernels with different sizes to construct the network framework. There are two main shortcomings of this approach: (1) the convolution operation based on multiple size convolution kernels requires enormous computation and has a low operational efficiency, which severely restricts its application in practical engineering. (2) The convolutional layer with a large size convolution kernel needs a mass of weight parameters, leading to a dramatic increase in the network training time and making it prone to overfitting in the case of small datasets. To address the above issues, a multi-scale dilated convolution network (MsDCN) is proposed for RUL prediction in this article. The MsDCN adopts a new multi-scale dilation convolution fusion unit (MsDCFU), in which the multi-scale network framework is composed of convolution operations with different dilated factors. This effectively expands the range of receptive field (RF) for the convolution kernel without an additional computational burden. Moreover, the MsDCFU employs the depthwise separable convolution (DSC) to further improve the operational efficiency of the prognostics model. Finally, the proposed method was validated with the accelerated degradation test data of rolling element bearings (REBs). The experimental results demonstrate that the proposed MSDCN has a higher RUL prediction accuracy compared to some typical CNNs and better operational efficiency than the existing multi-scale CNNs based on different convolution kernel sizes.

Detection of Cotton Plant Diseases Using Deep Transfer Learning

Agriculture is a vital part of every country’s economy, and India is regarded an agro-based nation. One of the main purposes of agriculture is to yield healthy crops without any disease. Cotton is a significant crop in India in relation to income. India is the world’s largest producer of cotton. Cotton crops are affected when leaves fall off early or become afflicted with diseases. Farmers and planting experts, on the other hand, have faced numerous concerns and ongoing agricultural obstacles for millennia, including much cotton disease. Because severe cotton disease can result in no grain harvest, a rapid, efficient, less expensive and reliable approach for detecting cotton illnesses is widely wanted in the agricultural information area. Deep learning method is used to solve the issue because it will perform exceptionally well in image processing and classification problems. The network was built using a combination of the benefits of both the ResNet pre-trained on ImageNet and the Xception component, and this technique outperforms other state-of-the-art techniques. Every convolution layer with in dense block is tiny, so each convolution kernel is still in charge of learning the tiniest details. The deep convolution neural networks for the detection of plant leaf diseases contemplate utilising a pre-trained model acquired from usual enormous datasets, and then applying it to a specific task educated with their own data. The experimental results show that for ResNet-50, a training accuracy of 0.95 and validation accuracy of 0.98 is obtained whereas training loss of 0.33 and validation loss of 0.5.