|
|
Ball Tracking in Sports Video
Motlík, Matúš ; Špaňhel, Jakub (referee) ; Bartl, Vojtěch (advisor)
This master's thesis deals with automatic detection and tracking of a soccer ball in sports videos. Based on the introduced techniques focusing on tracking of small objects in high-resolution videos, effective convolutional neural networks are designed and used by a modified version of tracking algorithm SORT for automatic object detection. A set of experiments with the processing of images in different resolutions and with various frequencies of detection extraction is carried out in order to examine the trade-off between processing speed and tracking accuracy. The obtained results of experiments are presented and used to form proposals for future work, which could lead to improvements in tracking accuracy while maintaining reasonable processing speed.
|
|
|
Holistic License Plate Recognition Based on Convolution Neural Networks
Le, Hoang Anh ; Hradiš, Michal (referee) ; Špaňhel, Jakub (advisor)
Main goal of this work was to create a holistic license plate reader, with an emphasis on achieving the highest possible accuracy on low quality images. Combination of convolutional and recurrent neural networks was designed and implemented, with usage of LSTM and CTC, where the inputs are cut-outs from the entire license plate. Competitive networks were also implemented to compare results. Networks were compared on a total of 4 datasets and the results were, that my design has achieved the best results with a recognition accuracy of 97.6%.
|
|
|
Smart Loudspeaker with Raspberry Pi
Vondráček, Tomáš ; Špaňhel, Jakub (referee) ; Herout, Adam (advisor)
The bachelor thesis deals with the creation of music system that allows users to organize and play music in real time. The music system is based on client-server architecture and consists of server, web and mobile application. The server application serves as a music player and a connection broker among clients. The client serves as a user interface for the server, but can organize and play music on its own. The music system is implemented in JavaScript (ECMAScript 2018). The server application is built on the Node.js and implements Socket.IO server with REST API. The web application is based on React and presented as SPA. The mobile application is implemented in React Native with focus on Android system. The server and web applications are deployed on Heroku servers and Raspberry Pi computer. Mobile application is published on Google Play. The music system can be used for private music playback or as a means of organizing music among multiple users.
|
|
|
Model Adaptation in Person Identification
Stratil, Jan ; Špaňhel, Jakub (referee) ; Hradiš, Michal (advisor)
This thesis deals with facial recognition using convolutional neural networks and with their current problems, which are pose, lighting and expression variance. It summarizes existing approaches, architectures and most recent loss functions. Further it deals with methods for rotating faces using GAN networks. In this thesis 3 neural networks are designed and trained for facial recognition. The best of them achieves 99.38% accuracy on LFW dataset and 88.08% accuracy on CPLFW dataset. Next face rotation network PCGAN is designed, which can be used for face frontalization or data augmentation purposes. This network is evaluated on Multi-PIE dataset and using the face frontalization it increases identification accuracy.
|
|
|
Web and Mobile App for Assigning and Confirming Tasks
Jeřábek, Filip ; Špaňhel, Jakub (referee) ; Herout, Adam (advisor)
This work solves managing employees by a dispatcher. The goal is to increase the efficiency of the work of a dispatcher, as well as a single employee's tasks. I focused on assigning tasks, event notification, and tasks state changes and all that in real-time. The solution consists of a website application that is used by a dispatcher and mobile application for employees. Website application is made in the PHP framework Symfony and the mobile app is made in the Flutter framework. The whole idea is to simplify the communication process. The dispatcher is able to check the state of assigned tasks whenever he needs, as well as the employee is notified about the newly assigned task, or changed task state, immediately. The solution significantly simplified the dispatcher's job. The idle times and unnecessary employee journeys were eliminated thanks to the immediate response, which made the whole process more effective. Following these facts, the overall costs should be lowered, and also more work should be made in the same amount of time.
|
|
|
Counting Vehicles in Image and Video
Gabzdyl, Dominik ; Herout, Adam (referee) ; Špaňhel, Jakub (advisor)
Analýza silničního provozu je stále náročnou úlohou. V průběhu této úlohy se vyskytují mnohá úskalí, která je třeba brát na vědomí. Například malé rozlišení obrazu, vysoký počet překrývajících se objektů, úhel kamery, rozmazání objektů v důsledku jejich pohybu nebo povětrnostní podmínky. Tato práce adresuje tato úskalí použitím konvolučních neuronových sítí. V této práci představuji novou architektu založenou na principu počítání regresí (Counting by Regression). Navržená architektura je inspirována některými state-of-the-art architekturami a vylepšuje přesnost na různých datasetech. Například na velmi malém PUCPR+ datasetu byla odmocnina ze střední kvadratické chyby (RMSE) snížena z 34.46 na 6.99 vozidel (měřeno na test setu). Dosažené výsledky ukázaly, že je zde stále prostor ke zlepšení a možný další výzkum v oblasti počítání regresí.
|
|
|
Detection of Traffic Signs and Lights
Oškera, Jan ; Špaňhel, Jakub (referee) ; Herout, Adam (advisor)
The thesis focuses on modern methods of traffic sign detection and traffic lights detection directly in traffic and with use of back analysis. The main subject is convolutional neural networks (CNN). The solution is using convolutional neural networks of YOLO type. The main goal of this thesis is to achieve the greatest possible optimization of speed and accuracy of models. Examines suitable datasets. A number of datasets are used for training and testing. These are composed of real and synthetic data sets. For training and testing, the data were preprocessed using the Yolo mark tool. The training of the model was carried out at a computer center belonging to the virtual organization MetaCentrum VO. Due to the quantifiable evaluation of the detector quality, a program was created statistically and graphically showing its success with use of ROC curve and evaluation protocol COCO. In this thesis I created a model that achieved a success average rate of up to 81 %. The thesis shows the best choice of threshold across versions, sizes and IoU. Extension for mobile phones in TensorFlow Lite and Flutter have also been created.
|
|
|
Vehicle On-Board Camera Analysis
Kadeřábek, Jan ; Bartl, Vojtěch (referee) ; Špaňhel, Jakub (advisor)
This thesis focuses on analysis of video from vehicle on-board camera. During the process of analysis, probihibitory traffic signs are detected and their specific type is classified. For recognized speed limit signs, their numeric value is extracted. From the processed information, it will try to create a file containing the unique occurrences of traffic signs including their GPS coordinates. For the purpose of detection and recognition of traffic signs, several data sets were created. A~cascade classifier with LBP features is used as a detector. Classification of the type and value of traffic signs is done using the k-Nearest Neigbour method.
|
|
|
Virtual Tour of FIT for Oculus Quest
Janů, Michal ; Špaňhel, Jakub (referee) ; Herout, Adam (advisor)
The main goal of this work is to make an application for VR headset Oculus Quest. This application has several features that allow the user to do more than just walk freely around the FIT BUT areal, such as Navigation, Instant travel, and viewing basic room information. The Navigation feature is used to find the shortest route to the desired office or lecture room, where every room has its panel with information about the room. The instant travel feature allowes choosing a starting location.
|
|
|
Web Application for Simple Management of Deadlines
Šmajzrová, Kateřina ; Špaňhel, Jakub (referee) ; Herout, Adam (advisor)
This work solves the creation of a simple web application that allows one administrator to easily manage his tasks, which he enters to his colleagues or students. The following text describes the design of the user interface and the database. I had solved this problem using the PHP framework Symfony, several scripts are written in Javascript, CSS framework Bootstrap was used to design the website. An important part of this work is user testing, data was obtained mainly through questionnaires, or personal consultation, and the results are in this thesis. After each testing phase, the existing version was improved based on the obtained data, bugs were fixed and the user interface was improved. The main benefit of the application is the simplification of assigning task to larger groups of people, it can be used for example by teachers or camp leaders. It will provide a quick overview of who has already fulfilled the tasks.
|
guest ::
RSS feed.