Deep neuralnetworks can suffer from the exploding and vanishing activation problem, in which the networks fail to train properly because the neural signals either amplify or attenuate across the layers and become sat...
详细信息
ISBN:
(数字)9783030863401
ISBN:
(纸本)9783030863401;9783030863395
Deep neuralnetworks can suffer from the exploding and vanishing activation problem, in which the networks fail to train properly because the neural signals either amplify or attenuate across the layers and become saturated. While other normalization methods aim to fix the stated problem, most of them have inference speed penalties in those applications that require running averages of the neural activations. Here we extend the unitary framework based on Lie algebra to neuralnetworks of any dimensionalities, overcoming the major constraints of the prior arts that limit synaptic weights to be square matrices. Our proposed unitary convolutional neuralnetworks deliver up to 32% faster inference speeds and up to 50% reduction in permanent hard disk space while maintaining competitive prediction accuracy.
The term Deep Learning or Deep neural Network refers to artificialneuralnetworks (ANN) with multi layers. Over the last few decades, it has been considered to be one of the most powerful tools, and has become very p...
详细信息
ISBN:
(纸本)9781538619490
The term Deep Learning or Deep neural Network refers to artificialneuralnetworks (ANN) with multi layers. Over the last few decades, it has been considered to be one of the most powerful tools, and has become very popular in the literature as it is able to handle a huge amount of data. The interest in having deeper hidden layers has recently begun to surpass classical methods performance in different fields;especially in pattern recognition. One of the most popular deep neuralnetworks is the Convolutional neural Network (CNN). It take this name from mathematical linear operation between matrixes called convolution. CNN have multiple layers;including convolutional layer, non-linearity layer, pooling layer and fully-connected layer. The convolutional and fully-connected layers have parameters but pooling and non-linearity layers don't have parameters. The CNN has an excellent performance in machine learning problems. Specially the applications that deal with image data, such as largest image classification data set (image Net), computer vision, and in natural language processing (NLP) and the results achieved were very amazing. In this paper we will explain and define all the elements and important issues related to CNN, and how these elements work. In addition, we will also state the parameters that effect CNN efficiency. This paper assumes that the readers have adequate knowledge about both machine learning and artificialneural network.
The technology using artificial intelligence (AI) allows, through the preprocessing of video and image data, their transformation into traffic information, the recognition in abusive conditions of elements, the specif...
详细信息
ISBN:
(纸本)9783031686498;9783031686504
The technology using artificial intelligence (AI) allows, through the preprocessing of video and image data, their transformation into traffic information, the recognition in abusive conditions of elements, the specification of the different characteristics of the vehicle (type, brand, model, color), object detection and tracking, automatic calibration and positioning. Algorithms for performance evaluation can evaluate recognition quality and detect malfunctions due to external factors, reduced lighting conditions, or incorrect camera installation. This article discusses the application of AI and deep learning techniques to identify vehicle wheels. It highlights the preprocessing of image and video data to extract traffic information, detect vehicle attributes, and perform object tracking. The study focuses on understanding AI and convolutional neuralnetworks (CNNs) for imageprocessing, emphasizing the significance of feature extraction algorithms. The research outlines the methodology for creating a dataset of annotated wheel images, selecting the YOLO detection algorithm, and using the Darknet Github framework for training and testing CNNs. It discusses the training process, including configuration, data augmentation, and evaluation metrics like Mean Average Precision (MAP) and Intersection Over Union (IOU). Finally, the article concludes with insights into the practical applications of the developed wheel detection system, suggesting future improvements and potential integrations with other projects.
The main objective of this project is to develop a technique to classify the ripeness of bananas into 3 categories, which is unripe, ripe and overripe systematically based on their histogram RGB value components. This...
详细信息
ISBN:
(纸本)9781424455614
The main objective of this project is to develop a technique to classify the ripeness of bananas into 3 categories, which is unripe, ripe and overripe systematically based on their histogram RGB value components. This system involved the process of collecting samples with different level of ripeness, imageprocessing and image classification by using artificialneural network. Collecting bananas sample is done by using Microsoft NX6000 webcam with 2 mega pixels. 32 samples were used as training samples for artificialneural network. In order to see whether the method mention above can classify the image correctly, another 28 images was used as a testing. From the result obtained, it was shown that the artificialneural network can generally classify the ripeness of bananas. This is because it can classify up to 25 samples correctly out of 28 samples. Developing a program totally by using Matlab version 7.0 can help classification process successfully.
We investigate the use of a Differential Vector Quantizer (DVQ) architecture for the coding of digital images. An artificialneural Network (ANN) is used to develop entropy-based codebooks which yield substantial data...
详细信息
ISBN:
(纸本)0819412015
We investigate the use of a Differential Vector Quantizer (DVQ) architecture for the coding of digital images. An artificialneural Network (ANN) is used to develop entropy-based codebooks which yield substantial data compression while retaining insensitivity to transmission channel errors. Two methods are presented for variable bit-rate coding using the described DVQ algorithm. In the first method, both the encoder and the decoder have multiple codebooks of different sizes. In the second, variable bit-rates are achieved by encoding using subsets of one fixed codebook. We compare the performance of these approaches under conditions of error-free and error-prone channels.
artificialneuralnetworks are applied to fiber-optic transmission. Two fiber-optic transmission techniques using neuralnetworks are proposed. One is an optical WDM (wavelength division multiplexing) demultiplexer co...
详细信息
artificialneuralnetworks are applied to fiber-optic transmission. Two fiber-optic transmission techniques using neuralnetworks are proposed. One is an optical WDM (wavelength division multiplexing) demultiplexer composed of a simple optical component and all electrical neural network. The other is a fiber-optic image-transmission technique using a multimode fiber and a neural network. In either technique, propagation modes, in an optical multimode guide, play an important role in the signal processing. Initial experimental results are presented for these techniques. The combination of optics and neuralnetworks have, so far, produced only the concept of optical neuralnetworks. The techniques described can be regarded as different approaches to this combination
The article presents a concept of the analysis of mechanical wear of prisms in the in-pavement airport lamps. The solution is based on imageprocessing technique that requires an appropriate selection of parameters du...
详细信息
ISBN:
(数字)9788362065424
ISBN:
(纸本)9788362065424
The article presents a concept of the analysis of mechanical wear of prisms in the in-pavement airport lamps. The solution is based on imageprocessing technique that requires an appropriate selection of parameters due to the specificity of the objects. During the experimental tests, a database consisting of 316 photos of IDM airport lamps mounted in the airport areas was used. The proposed solution using an artificialneural network allows for the classification of lamps with an efficiency of 81.4%.
Multispectral imagery is a large domain with a number of practical applications: thermography and quality control in industry, food science, and agronomy. The main interest is to obtain spectral information of the obj...
详细信息
Multispectral imagery is a large domain with a number of practical applications: thermography and quality control in industry, food science, and agronomy. The main interest is to obtain spectral information of the objects for which a reflectance signal can be associated to physical, chemical, and/or biological properties. Agronomic applications of multispectral imagery generally involve the acquisition of several images in visible and near infrared wavelengths. We first present a different kind of multispectral devices used for agronomic applications and then introduce an original multispectral acquisition system based on a single CCD. First results in laboratory are detailed, presenting a detection method using a neural network and in-field acquisitions and their results are shown. To improve the quality of weed detection, the spatial distribution of crops is used by a second method. Finally, the first works on merging are outlined. (C) 2004 SPIE and IST.
作者：
Alwis, SAustin, JUniv York
Dept Comp Sci Adv Comp Architecture Grp York YO10 5DD N Yorkshire England
This paper describes a novel massively parallel connectionist architecture for image retrieval. The proposed search engine of the system consists of associative memory nodes connected by information channels which con...
详细信息
ISBN:
(纸本)3540660682
This paper describes a novel massively parallel connectionist architecture for image retrieval. The proposed search engine of the system consists of associative memory nodes connected by information channels which convey symbolic messages. Symbolic information stored inside the system is obtained using gestalt feature extraction methods which capture multiple representations of images. In this paper, we summarise our feature extraction method and then we describe the connection schemata of the system, training process as well as how such a system can be utilised to capture perceptual similarity of trademark images. Finally, we present results obtained during evaluation of the system.
For the automatic inspection for printed labels, which are covered with rubber-like coatings and Curl, we have developed a camera-based portable inspection system. In this paper, we explained the developed system, and...
详细信息
ISBN:
(纸本)9783642024801
For the automatic inspection for printed labels, which are covered with rubber-like coatings and Curl, we have developed a camera-based portable inspection system. In this paper, we explained the developed system, and especially discuss the inspection method of the spread and chip of the printed labels using neuralnetworks. The experimental results confirm the validity of the proposed method for the spread and chip of alphanumerics.
暂无评论