|
|
Vehicle Counting in Still Image
Jelínek, Zdeněk ; Juránek, Roman (referee) ; Špaňhel, Jakub (advisor)
The main goal of this thesis was to compare different approaches to vehicle counting by density estimation. Four convolutional neural networks were tested - Counting CNN, Hydra CNN, Perspective-Aware CNN and Multi-column CNN. The evaluation of these models was done on three different datasets. The Perspective-aware CNN has achieved the most accurate results across all datasets. This model has reached 2.86 Mean Absolute Error on the PUCPR+ dataset, proving that it is the most suitable for the vehicle counting problem.
|
|
|
The use of magnetic microparticles for DNA isolation
Jelínek, Zdeněk ; Horák, Daniel (referee) ; Rittich, Bohuslav (advisor)
The effectiveness of magnetic microparticles in isolation of DNA from Lactobacillus rhamnosus CCM 1825T and DNA from chicken erythrocytes were studied in diploma thesis. Magnetic HEMA based microparticles coated by carboxylic groups and hyperbranched styrene-divinylbenzene particles (IMC AS ČR, Prague, Czech Republic) were used for DNA isolation. Magnetic microparticles Dynabeads® DNA DIRECT™ Universal (Dynal, Norway) based on polystyrene and MPG® Uncoated (PureBiotech, USA) based on magnetic glass were used as a control. The dependence of amount of eluted DNA on concentration of DNA in the base solution and the dependence of amount of eluted DNA on concentration of magnetic microparticles were studied. The affinity of magnetic microparticles to RNA for various concentrations of RNA solution was studied, too. The ability of tested particles to isolate DNA from real samples was validated using milk product Actimel. The quality of isolated DNA of Lactobacillus genus was proved using genus specific PCR.
|
|
|
Molecular typisation of lactic acid bacteria
Jelínek, Zdeněk ; Ing. Špano Miroslav,PhD. (referee) ; Rittich, Bohuslav (advisor)
Lactic acid bacteria play important role both in food industry and health care. They are part of many fermented dairy products and gastrointestinal microflora. There is a whole variety of methods for bacteria identification, but we will focus on Rep-PCR and RAPD only. The aim of this work will be to discuss the impact of different correlation coefficients and cluster analysis algorithms on the results of taxonomic analysis.
|
|
|
Company liquidation and its risks
Hedbávný, Ondřej ; Jelínek, Zdeněk (referee) ; Mašterová, Lucie (advisor)
The aim of this report is company liquidation and her possible risks.The theoretical part of this dissertation deal with the general law business liquidation of the company, primarily focused on the legal status of the liquidator and his individual acts within the liquidation process.The analytical part analyse the course of liquidation of a particular company from the perspective of liquidator. The target is carry out problem-free liquidation with the greatest possible benefit for shareholders.
|
|
| |
|
|
Planning and Drawing Module for Matches
Jelínek, Zdeněk ; Soukup, Ondřej (referee) ; Křivka, Zbyněk (advisor)
This bachelors thesis consists of description of a design and implementation of a module for drawing tournaments in competitive sports. The goal of the module is to match competitors in such a way that the repetitions of matches are minimal and the quality of the individual matches is the best, and that each competitor plays appropriate number of matches. The worst-case asymptotic time complexity of the resulting drawing algorithm is polynomial.
|
|
|
Vehicle Re-Identification Using Vision Transformers
Jelínek, Zdeněk ; Hradiš, Michal (referee) ; Špaňhel, Jakub (advisor)
The main objective of this thesis was to investigate the potential of vision transformers in vehicle re-identification. Convolutional neural networks have so far dominated this area of computer vision. In total, two models have been tested - TransReID and CMT. TransReID is a model based purely on vision transformers and was created specifically for vehicle re-identification. The main part of the experiments with this model was devoted to the use of key points on the vehicle. With proper extraction of the regions around the key points and the use of post-processing, I achieved state-of-the-art results. The CMT model is a combination of convolutional networks and transformers that was not designed for vehicle re-identification. I modified the model and conducted extensive experiments with it to obtain the best configuration for vehicle re-identification. I evaluated the models on the standard datasets VeRi-776, VehicleID, CityFlowV2-ReID and CarsReId74k and compared with state-of-the-art models. With the CMT model, I achieved the best result of 0.860 on the mAP metric on the VeRi-776 dataset and the best result of 97.6% on the Rank5 metric on the VehicleID dataset.
|
|
|
Vehicle Counting in Still Image
Jelínek, Zdeněk ; Juránek, Roman (referee) ; Špaňhel, Jakub (advisor)
The main goal of this thesis was to compare different approaches to vehicle counting by density estimation. Four convolutional neural networks were tested - Counting CNN, Hydra CNN, Perspective-Aware CNN and Multi-column CNN. The evaluation of these models was done on three different datasets. The Perspective-aware CNN has achieved the most accurate results across all datasets. This model has reached 2.86 Mean Absolute Error on the PUCPR+ dataset, proving that it is the most suitable for the vehicle counting problem.
|
|
| |
|
| |
guest ::
