|
|
Self-supervised learning in computer vision applications
Vančo, Timotej ; Richter, Miloslav (referee) ; Janáková, Ilona (advisor)
The aim of the diploma thesis is to make research of the self-supervised learning in computer vision applications, then to choose a suitable test task with an extensive data set, apply self-supervised methods and evaluate. The theoretical part of the work is focused on the description of methods in computer vision, a detailed description of neural and convolution networks and an extensive explanation and division of self-supervised methods. Conclusion of the theoretical part is devoted to practical applications of the Self-supervised methods in practice. The practical part of the diploma thesis deals with the description of the creation of code for working with datasets and the application of the SSL methods Rotation, SimCLR, MoCo and BYOL in the role of classification and semantic segmentation. Each application of the method is explained in detail and evaluated for various parameters on the large STL10 dataset. Subsequently, the success of the methods is evaluated for different datasets and the limiting conditions in the classification task are named. The practical part concludes with the application of SSL methods for pre-training the encoder in the application of semantic segmentation with the Cityscapes dataset.
|
|
|
Object detection in video using neural networks
Mikulský, Petr ; Sikora, Pavel (referee) ; Myška, Vojtěch (advisor)
This diploma thesis deals with the detection of moving objects in a video recording using neural networks. The aim of the thesis was to detect road users in video recordings. Pre-trained YOLOv5 object detection model was used for a practical part of the thesis. As part of the solution, an own dataset of traffic road video recordings was created and annotated with following classes: a car, a bus, a van, a motorcycle, a truck and a trailer truck. Final version of this dataset comprise 5404 frames and 6467 annotated objects in total. After training, the YOLOv5 model achieved 0.995 mAP, 0.995 precision and 0.986 recall on the dataset. All steps leading to the final form of the dataset are described in the conclusion chapter.
|
|
|
Low-Dimensional Matrix Factorization in End-To-End Speech Recognition Systems
Gajdár, Matúš ; Grézl, František (referee) ; Karafiát, Martin (advisor)
The project covers automatic speech recognition with neural network training using low-dimensional matrix factorization. We are describing time delay neural networks with factorization (TDNN-F) and without it (TDNN) in Pytorch language. We are comparing the implementation between Pytorch and Kaldi toolkit, where we achieve similar results during experiments with various network architectures. The last chapter describes the impact of a low-dimensional matrix factorization on End-to-End speech recognition systems and also a modification of the system with TDNN(-F) networks. Using specific network settings, we were able to achieve better results with systems using factorization. Additionally, we reduced the complexity of training by decreasing network parameters with the use of TDNN(-F) networks.
|
|
|
Document Quality Enhancement
Trčka, Jan ; Zemčík, Pavel (referee) ; Juránek, Roman (advisor)
The aim of this work is to increase the accuracy of the transcription of text documents. This work is mainly focused on texts printed on degraded materials such as newspapers or old books. To solve this problem, the current method and problems associated with text recognition are analyzed. Based on the acquired knowledge, the implemented method based on GAN network architecture is chosen. Experiments are a performer on these networks in order to find their appropriate size and their learning parameters. Subsequently, testing is performed to compare different learning methods and compare their results. Both training and testing is a performer on an artificial data set. Using implemented trained networks increases the transcription accuracy from 65.61 % for the raw damaged text lines to 93.23 % for lines processed by this network.
|
|
|
Deep Book Recommendation
Gráca, Martin ; Beran, Vítězslav (referee) ; Hradiš, Michal (advisor)
This thesis deals with the field of Recommendation systems using Deep Neural Networks and their use in book recommendation. There are the main traditional recommender systems analysed and their representations are summarized, as well as systems with more advancec techniques based on machine learning.. The core of the thesis is the use of convolutional neural networks for natural language processing and the creation of a book recommendation system. Suggested system make recommendation based on user data, including user reviews and book data, including full texts.
|
|
|
Identification of vertebrae type in CT data by machine learning methods
Matoušková, Barbora ; Kolář, Radim (referee) ; Chmelík, Jiří (advisor)
Identification of vertebrae type by machine learning is an important task to facilitate the work of medical doctors. This task is embarrassed by many factors. First, a spinal CT imagining is usually performed on patiens with pathologies such as lesions, tumors, kyphosis, lordosis, scoliosis or patients with various implants that cause artifacts in the images. Furthermore, the neighboring vertebraes are very similar which also complicates this task. This paper deals with already segmented vertebrae classification into cervical, thoracic and lumbar groups. Support vector machines (SVM) and convolutional neural networks (CNN) AlexNet and VGG16 are used for classification. The results are compared in the conclusion.
|
|
|
Klasifikace dat v obraze pomocí nástrojů pro strojové učení v jazyce Python
Voronin, Artyom ; Appel, Martin (referee) ; Bastl, Michal (advisor)
This thesis introduces the issue of data classification in the image using tools for machine learning in Python. The aim is to verify the possibilities of overtraining existing models on their own data and evaluating the efficiency and complexity of the entire process. Subsequently, the processing of the achieved results in the form of a demonstration task, image capturing by a web camera and classification of the object in the field of view.
|
|
|
Vehicle Control via Reinforcement Learning
Maslowski, Petr ; Uhlíř, Václav (referee) ; Šůstek, Martin (advisor)
The goal of this thesis is a creation of an autonomous agent that can control a vehicle. The agent utilizes reinforcement learning that uses neural networks. The agent interprets images from the front vehicle camera and selects appropriate actions to control the vehicle. I designed and created reward functions and then experimented with hyperparameters setup. Trained agent simulate driving on the road. The result of this thesis shows a possible approach to control an autonomous vehicle agent using machine learning method in CARLA simulator.
|
|
|
Deep Book Recommendation
Gráca, Martin ; Kolář, Martin (referee) ; Hradiš, Michal (advisor)
This thesis deals with the field of recommendation systems using deep neural networks and their use in book recommendation. There are the main traditional recommender systems analysed and their representations are summarized, as well as systems with more advanced techniques based on machine learning. The core of the thesis is to use convolutional neural networks for natural language processing and create a hybrid book recommendation system. Suggested system includes matrix factorization and make recommendation based on user ratings and book metadata, including texts descriptions. I designed two models, one with bag-of-words technique and one with convolutional neural network. Both of them defeat baseline methods. On the created data set, that was created from the Goodreads, model with CNN beats model with BOW.
|
|
|
Reinforcement Learning for RoboCup
Bočán, Hynek ; Škoda, Petr (referee) ; Smrž, Pavel (advisor)
Goal of this thesis is creation of artificial intelligence capable of controlling robotic soccer player simulated in SimSpark environment. Agent created is expanding capabilities of existing third party agent which provides set of basic skills such as localization on the field, dribbling with the ball and omnidirectional walk. Responsibility of the created agent is to pick the best action based current state of the game. This decision making was implemented using reinforcement learning and its method Q-learning. State of the game is transformed into 2D picture with several planes. This picture is then analyzed using deep convolution neural network implemented using C++ and DeepCL library.
|
guest ::
