|
|
Measurement of Reaction Time and Attention
Oravová, Pavlína ; Provazník, Ivo (referee) ; Sekora, Jiří (advisor)
The semesters thesis is focused on Measurement of Reaction Time and Attention. Part of this work is a definition of reaction time and factors influencing reaction time. Measuring reaction time in context of the theory of intelligence C-H-C is described together with this theory. There are included related tests, specifically Stroop test, Flanker test, Go/No-go test and N-back test. Next there is mentioned online problematics containing hardware and software delays and diagnostic and measuring of the network. The last part of the work is dedicated to a practical part.
|
|
|
Emotional Cartography
Rygálová, Monika ; Pfeiffer, Jan (referee) ; Sterec, Pavel (advisor)
The work aims to artistically compile perception of maps as a fact, what let us think about The world - how we know it from the map. I work with data, which I gain by displaying technology – eye tracking, which helps me to record track of moving eyes during watching any kind of picture. The observations will be people from different places such a place of stay etc. Gained data I am going to use as a study, which I will componate to maps and different vizualizations of world, countries etc. "Where i have not ever been before, it does not exist" – is idea of percepting world, wich I also work with on that project. Study will contain all aspect of perceiving maps and systematicly shown world related to the person, his memories to that place, fyzical contact and his impact in his scale person versus a the world.
|
|
| |
|
|
Analysis of driver’s conduct during solving of situations associated with pedestrians crossing the road
Maxera, Pavel ; Kolíbal, Zdeněk (referee) ; Rábek, Vlastimil (referee) ; Kledus, Robert (advisor)
The doctoral thesis analyses driver’s conduct while solving situations associated with crossing of pedestrians across the road in cases of various design of pedestrian crossings and at different conditions. The thesis deepens the knowledge of the human factor impact on the occurrence of a traffic accident involved vehicle and pedestrian and thesis also complements knowledge for the needs of the analysis of traffic accidents, especially in solving the pre collision phase and at assessment of possibilities for collision prevention by involved participants. The thesis deals with driver’s conduct, various models of the conduct as well as the thesis focuses on the visual perception, the process of information processing, the driver’s conduct and the reaction time. In terms of the solution suitable types of experiments were designed and implemented. Based on performed measurements a method of processing and evaluating data on drivers’ conduct was found as well as more significant data set was obtained for a detailed analysis of drivers' conduct in different driving situations. The assessed quantities of drivers' conduct were analysed with respect to the dangerousness of driving situations. For these purposes, the categories of the dangerousness of driving situations were defined (situations completely safe, with increased danger, dangerous and critical), into which the analysed driving situations were subsequently included. To enable the quantification of this classification of situations into the categories of the dangerousness, the coefficient of the dangerousness (so called K) was defined. From the detailed analysis of the obtained data, the limit values of this coefficient were determined, and these were subsequently verified using data from the solution of real traffic accidents. Concurrently the analysis verified the suitability of using this hazard coefficient in the analysis of traffic accidents, especially for a detailed assessment of the possibilities of collision prevention.
|
|
|
Checking of the Situation Behind the Vehicle
Perničková, Tereza ; Bucsuházy, Kateřina (referee) ; Belák, Michal (advisor)
This diploma thesis deals with the control of the situation behind the vehicle. The theoretical part of the thesis summarizes the knowledge about the transport system, the reaction time and the individual phases, including the factors influencing the reaction time and the possible methods of its measurement. Further, there are issues of vision from the vehicle and the rules of safe driving. Explanation is the eyetracking method, the types of these devices, and the methods of visualizing the data obtained from the records. In the analytical part are drawn from data recorded video. There are testing route, testing device, testing vehicle and information about testing drivers. The frequency and length of the individual components of the mirror views were evaluated depending on the selected mirror even without dependence from the data obtained.
|
|
|
Transformer Neural Networks for Handwritten Text Recognition
Vešelíny, Peter ; Beneš, Karel (referee) ; Kohút, Jan (advisor)
This Master's thesis aims to design a system using the transformer neural network and perform experiments with this proposed model in the task of handwriting text recognition. In this thesis, a multilingual dataset with predominate Czech texts is used. The experiments examine the influence of basic hyperparameters, such as network size, convolutional encoder type, and the use of different text tokenizers. In this work, I also use text corpora of the Czech language which is used to train the network decoder. Furthermore, I experiment with the usage of additional textual information during the decoding process. This information comes from the previous line of the transcribed image. The transformer achieves a character recognition error rate of 3.41 % on the test data set which is 0.16 % worse performance than the recurrent neural network achieves. To compare this model with other transformer-based models from available articles, the network was trained on the IAM dataset, where it achieved an error of 2.48 % and therefore outperformed other models in handwriting text recognition task.
|
|
|
Convolutional Networks for Historic Text Recognition
Vešelíny, Peter ; Kolář, Martin (referee) ; Kišš, Martin (advisor)
This thesis deals with text line recognition of historical documents. Historical texts dating back to the 17th - 19th centuries are written in fraktur typeface. The character recognition problem is solved using neural network architecture called sequence-to-sequence . This architecture is based on encoder-decoder model and contains attention mechanism. In this thesis a dataset, from texts originated from German archiv called Deutsches Textarchiv , was created. This archive contains 3 897 different German books that have available transcripts and corresponding images of pages. The created dataset was used to train and experiment with the proposed neural network. During the experiments, several convolutional models, hyperparameters and the effects of positional embedding were investigated. The final tool can recognize characters with accuracy 99,63 %. The contribution of this work is the~mentioned dataset and neural network, which can be used to recognize historical documents.
|
|
|
Vliv pozornosti investora na trh s ropou
Topolnikova, Anna
Topolnikova, A. The influence of investor attention on the oil market. Bachelor thesis. Brno: Mendel University in Brno, 2023. The aim of this bachelor thesis is to identify the impact of selected indicators on the profitability of oil futures contracts and the profitability of shares of companies operating in the field of oil production. The theoretical part of the thesis describes the possibilities of investing in oil, presents the most important methods of stock valuation and determinants affecting oil prices. A separate chapter is devoted to investor sentiment. Where methods of measuring sentiment and expressing attention are presented with the help of searching for a certain term in Google Trends. The practical part is devoted to the relationship between the performance of ExxonMobil stock, WTI crude oil futures contracts and other selected indicators through correlation and regression analysis. Based on the information obtained from the theoretical and practical parts of the work, recommendations for investors are formulated.
|
|
|
Relationship between hand grip strength and cognitive function in older adults over 65 years of age
Čutková, Michaela ; Šteffl, Michal (advisor) ; Hráský, Pavel (referee)
Title: Relationship between hand grip strength and cognitive function in older adults over 65 years of age Goals: The main goal of this diploma thesis was to find out possible links between hand grip strength and cognitive functions in older adults over 65 years of age Method: As part of the master's thesis, a secondary analysis of data from the international cross- sectional study conducted within the Survey of Health Ageing and Retirement in Europe (SHARE) project was performed. The participants of the study were divided into seven categories based on their educational attainment, separately for both men and women. Generalized linear models were calculated for each of the mentioned categories as part of the data analysis. Results: The survey included a total of 38,519 participants with an average age of 74.00 ± 6.7 years from 18 countries worldwide. More than half of the participants were women. In the data analysis, it was found that the most represented educational group among the participants was the third group with higher secondary education, as defined by the ISCED 2011 classification. The tables indicate that values of all percentiles increase linearly with higher levels of education. This result suggests that higher education is associated with better performance in the examined test....
|
|
|
Deep Neural Networks for Historical Document Classification
Pinkeová, Bettina ; Kohút, Jan (referee) ; Kišš, Martin (advisor)
The aim of this work is to create a system for historical documents classification . The task is specifically about classification of documents according to the place of origin. Several systems are proposed for solving this problem, in the work. The first designed and implemented system is based on a convolutional neural network with a self-attention mechanism instead of an average pooling layer. Another system is based on the BEiT model, which is built on a visual transformer. The BEiT model was pretrained on the task of masked image modelling and subsequently trained on the given classification task. The system based on convolutional neural network achieved an accuracy of 81.6% and the system based on masked image modelling achieved an accuracy of 82.9%. The systems implemented in this work, surpassed the systems participating in the ICDAR 2021 conference in terms of success.
|
guest ::
