|
|
Deep neural networks and their application for image data processing
Golovizin, Andrey ; Mrázová, Iveta (advisor) ; Holan, Tomáš (referee)
In the area of image recognition, the so-called deep neural networks belong to the most promising models these days. They often achieve considerably better results than traditional techniques even without the necessity of any excessive task-oriented preprocessing. This thesis is devoted to the study and analysis of three basic variants of deep neural networks-namely the neocognitron, convolutional neural networks, and deep belief networks. Based on extensive testing of the described models on the standard task of handwritten digit recognition, the convolutional neural networks seem to be most suitable for the recognition of general image data. Therefore, we have used them also to classify images from two very large data sets-CIFAR-10 and ImageNet. In order to optimize the architecture of the applied networks, we have proposed a new pruning algorithm based on the Principal Component Analysis. Powered by TCPDF (www.tcpdf.org)
|
|
|
Deep neural networks and their implementation
Vojt, Ján ; Mrázová, Iveta (advisor) ; Božovský, Petr (referee)
Deep neural networks represent an effective and universal model capable of solving a wide variety of tasks. This thesis is focused on three different types of deep neural networks - the multilayer perceptron, the convolutional neural network, and the deep belief network. All of the discussed network models are implemented on parallel hardware, and thoroughly tested for various choices of the network architecture and its parameters. The implemented system is accompanied by a detailed documentation of the architectural decisions and proposed optimizations. The efficiency of the implemented framework is confirmed by the results of the performed tests. A significant part of this thesis represents also additional testing of other existing frameworks which support deep neural networks. This comparison indicates superior performance to the tested rival frameworks of multilayer perceptrons and convolutional neural networks. The deep belief network implementation performs slightly better for RBM layers with up to 1000 hidden neurons, but has a noticeably inferior performance for more robust RBM layers when compared to the tested rival framework. Powered by TCPDF (www.tcpdf.org)
|
|
|
Convolutional neural networks and their application in object detection
Hrinčár, Matej ; Mrázová, Iveta (advisor) ; Pešková, Klára (referee)
1 Title: Convolutional neural networks and their application in object detection Author: Matej Hrinčár Department: Department of Theoretical Computer Science and Mathematical Logic Supervisor: doc. RNDr. Iveta Mrázová, CSc. Supervisor's e-mail address: Iveta.Mrazova@mff.cuni.cz Abstract: Nowadays, it has become popular to enhance live sport streams with an augmented reality like adding various statistics over the hockey players. To do so, players must be automatically detected first. This thesis deals with such a challenging task. Our aim is to deliver not only a sufficient accuracy but also a speed because we should be able to make the detection in real time. We use one of the newer model of neural network which is a convolutional network. This model is suitable for proces- sing image data a can use input image without any preprocessing whatsoever. After our detailed analysis we choose this model as a detector for hockey players. We have tested several different architectures of the networks which we then compared and choose the one which is not only accurate but also fast enough. We have also tested the robustness of the network with noisy patterns. Finally we assigned detected pla- yers to their corresponding teams utilizing K-mean algorithm using the information about their jersey color. Keywords:...
|
|
| |
|
|
Convolutional neural networks and their implementation
Schmid, Martin ; Mrázová, Iveta (advisor) ; Petříčková, Zuzana (referee)
Bachelor thesis describes using convolutional neural networks for recognizing symbols from images. First describes this model and shows it's implementation. Then this implementation is used for sample application. First, model of neural networks is described, then learning of this model (including backpropagation algorithm). Finally, convolutional neural networks are presented with it's advantages for symbol recognition. Then some existing implementations of neural networks are analyzed, including speed comparison. None of these implementations support convolutional networks, so this model is added to one of them. Then this extension and it's interface (how to use it) is presented. To show features of this model and to prove functionality of the implementation, sample application is created. This application is available on the web site and runnable using only a web browser. Keywords: Convolutional neural networks, OCR, Encog 7
|
|
|
Convolutional Neural Networks for Emotion Recognition
Jileček, Jan ; Najman, Pavel (referee) ; Hradiš, Michal (advisor)
Convolutional neural networks are used for various tasks, but foremost in machine learning, in which they excel. This work is going to introduce some existing frameworks, other algorithms for recognition and then we describe the training dataset creation and the model for emotion recognition training process. Mentioned model has accuracy of 60%. It is used for emotion statistics retrieval from movie trailers. Model for genre recognition is created from those statistics and then finally used in our application for genre recognition of the input trailer, with best accuracy of 47%.
|
|
|
Convolutional Neural Networks for Security Applications
Kišš, Martin ; Hradiš, Michal (referee) ; Smrž, Pavel (advisor)
This thesis deals with design and implementation of application for person recognition in security camera. For single face rocongition are used convolutional neural networks, which creates representation of the face, and k-nearest neighbours algorithm for classification. For recognition of sequence of faces there are three algorithms implemented. On test data success of recognition reached nearly 75 %.
|
|
|
Learning to Generate Images with Convolutional Neural Networks
Kohút, Jan ; Kolář, Martin (referee) ; Hradiš, Michal (advisor)
The aim of this Bachelor's thesis is to design and analyze convolutional neural networks generating images of characters based on their parameters. Parameters of characters are type of char, font, colour of character, background colour, translation and rotation. Neural networks have created multidimensional representation of each parameter. Relations inside these representation are similar to relations inside parameters. Neural networks generate characters with new values of parameters based on interpolation between learned values of parameters. Neural networks are capable to generalize problem of generating images.
|
|
|
Depth Estimation by Convolutional Neural Networks
Ivanecký, Ján ; Španěl, Michal (referee) ; Hradiš, Michal (advisor)
This thesis deals with depth estimation using convolutional neural networks. I propose a three-part model as a solution to this problem. The model contains a global context network which estimates coarse depth structure of the scene, a gradient network which estimates depth gradients and a refining network which utilizes the outputs of previous two networks to produce the final depth map. Additionally, I present a normalized loss function for training neural networks. Applying normalized loss function results in better estimates of the scene's relative depth structure, however it results in a loss of information about the absolute scale of the scene.
|
|
|
Image Captioning with Recurrent Neural Networks
Kvita, Jakub ; Španěl, Michal (referee) ; Hradiš, Michal (advisor)
Tato práce se zabývá automatickým generovaním popisů obrázků s využitím několika druhů neuronových sítí. Práce je založena na článcích z MS COCO Captioning Challenge 2015 a znakových jazykových modelech, popularizovaných A. Karpathym. Navržený model je kombinací konvoluční a rekurentní neuronové sítě s architekturou kodér--dekodér. Vektor reprezentující zakódovaný obrázek je předáván jazykovému modelu jako hodnoty paměti LSTM vrstev v síti. Práce zkoumá, na jaké úrovni je model s takto jednoduchou architekturou schopen popisovat obrázky a jak si stojí v porovnání s ostatními současnými modely. Jedním ze závěrů práce je, že navržená architektura není dostatečná pro jakýkoli popis obrázků.
|
guest ::
