|
|
Monitoring Pedestrian by Drone
Dušek, Vladimír ; Goldmann, Tomáš (referee) ; Drahanský, Martin (advisor)
This thesis is focused on monitoring people in a video footage captured by drone. People are detected by trained model of detector RetinaNet. A feature vector is extracted for each detected person using color histograms. Identification of people is realized by comparing their feature vectors with respect to their distance in the frame. In the end the trajectories of all people are visualized in a panorama image. Accuracy of the trained RetinaNet detector on difficult validation data is 58.6 %. Error rate is partially reduced by the way of algorithm design for trajectory visualisation. It's not necessary to successfully detect person on every frame for correct visualization of its trajectories. At the same time, static objects which are detected as person but are not moving are not consider as people and are not visualized at all. There is a lot of algorithms dealing with people detection however only a few approaches are focused on detection people from an aerial footage.
|
|
|
Implementation of Deep Learning Algorithm on Embedded Device
Ondrášek, David ; Boštík, Ondřej (referee) ; Horák, Karel (advisor)
This thesis deals with the implementation of inference model, based on the methods of deep learning, on embedded device. First, machine learning and deep learning methods are researched with emphasis on state-of-the-art techniques. Next, the best suitable hardware had to be selected. In the conclusion, two devices are chosen: Jetson Nano and Raspberry Pi. Then the custom dataset, consisting of three classes of candies, was created and used for training custom inference model through the transfer learning technique. Model is later used in the application, capable of object detection. Application is implemented on Jetson Nano and Raspberry Pi and then evaluated.
|
|
|
Segmentation of cartilage tissue of mouse embryos in 3D micro CT data
Matula, Jan ; Vičar, Tomáš (referee) ; Chmelík, Jiří (advisor)
Manual segmentation of cartilage tissue in micro CT images of mouse embryos is a very time consuming process and significantly increases the time required for the research of mammal facial structure development. This problem might be solved by using a fully-automatic segmentation algorithm. In this diploma thesis a fully-automatic segmentation method is proposed using a convolutional neural network trained on manually segmented data. The architecture of the proposed convolutional network is based on the U-Net architecture with it's encoding part substituted for the encoding part of the VGG16 classification convolutional neural network pretrained on the ImageNet database of labeled images. The proposed network achieves Dice coefficient 0.8731 ± 0.0326 in comparison to manually segmented images.
|
|
|
Deep learning methods for vessel and optic disc segmentation in ophthalmologic sequences
Rozhoňová, Andrea ; Odstrčilík, Jan (referee) ; Hesko, Branislav (advisor)
The aim of the following thesis was to study the issue of optical disc and retinal vessels segmentation in ophthalmologic sequences. The theoretical part of the thesis summarizes the principles of different approaches in the field of deep learning, which are used in connection with the given issue. Based on the theoretical part, methods for optical disk segmentation and retinal vessel segmentation based on the convolutional neural networks Linknet, PSPNet, Unet and MaskRCNN are proposed. The practical part of the thesis deals with the description of their implementation and subsequent evaluation.
|
|
|
Multiclass segmentation of 3D medical data using deep learning
Slunský, Tomáš ; Uher, Václav (referee) ; Kolařík, Martin (advisor)
Master's thesis deals with multiclass image segmentation using convolutional neural networks. The theoretical part of the Master's thesis focuses on image segmentation. There are basics principles of neural networks and image segmentation with more types of approaches. In practical part the Unet architecture is choosen and is described for image segmentation more. U-net was applied for medicine dataset. There is processing procedure which is more described for image proccesing of three-dimmensional data. There are also methods for data preproccessing which were applied for image multiclass segmentation. Final part of current master's thesis evaluates results.
|
|
|
Detection of Vehicle License Plates in Video
Líbal, Tomáš ; Hradiš, Michal (referee) ; Herout, Adam (advisor)
This thesis deals with preparation of training dataset and training of convolutional neural network for licence plate detection in video. Darknet technology was used for detection, specifically the YOLOv3-tiny neural network model. The solution was focused on the most accurate detection and the smallest number of false positives per image, thus minimizing overall model error. Dataset was prepared from existing freely available datasets, from the dataset provided by the GRAPH@FIT research group, and from self-annotated images created from downloaded YouTube videos. Furthermore, this dataset has been processed using data augmentation, extending it to twice the size. The YOLO Mark tool was used to create annotations. An ROC curve was used to visualize the detection success. Created solution reaches minimum total error 10,849%. Part of the solution is already mentioned dataset.
|
|
|
Evolutionary Algorithms for Neural Networks Learning
Vosol, David ; Rozman, Jaroslav (referee) ; Zbořil, František (advisor)
Main point of this thesis is to find and compare posibilities of cooperation between evolutionary algorithms and neural network learning and their comparison with classical learning technique called backpropagation. This comparison is demonstrated with deep feed-forward neural network which is used for classification tasks. The process of optimalization is via search of optimal values of weights and biases within neural network with fixed topology. We chose three evolutionary approaches. Genetic algorithm, differential evolution and particle swarm optimization algorithm. These three approaches are also compared between each other. The demonstrating program is implemented in Python3 programming language without usage of any third parties libraries focused on deep learning.
|
|
|
Neural network for style transfer
Kadlec, Filip ; Matoušek, Radomil (referee) ; Hůlka, Tomáš (advisor)
In this bachelor’s thesis we describe machine learning, types of artificial neural networks and internal processes of neural networks, such as feedforward data processing and training neural networks. We are also pursuing comparison and description of libraries (such as TensorFlow and Keras), which are suitable for neural networks implementation. In the practical part of thesis, we are dealing with problem called artistic style transfer with convolutional neural network.
|
|
|
Recommender System for Web Articles
Kočí, Jan ; Kesiraju, Santosh (referee) ; Fajčík, Martin (advisor)
Tématem této bakalářské práce jsou doporučovací systémy pro webové články. Tato práce nejdříve uvádí nejpopulárnější metody z této oblasti a vysvětluje jejich principy, následně navrhuje požití vlastní architektury, založené na neuronových sítích, která aplikuje metodu Skip-gram negative sampling na problematiku doporučování. V další části pak implementuje tuto architekturu společně s několika dalšími modely, požívající algoritmus SVD, collaborative filtering s algoritmem ALS a také metodu Doc2Vec k vytvoření vektorové reprezentace z obsahu získaných článků. Na závěr vytváří tři evaluační metriky, konkrétně metriky RANK, Recall at k a Precision at k, a vyhodnocuje kvalitu implementovaných modelů srovnáním výsledků s nejmodernějšími modely. Kromě toho také diskutuje o roli a smyslu doporučovacích systémů ve společnosti a uvádí motivaci pro jejich používání.
|
|
|
Melody Extraction with Deep Learning
Balhar, Jiří ; Hajič, Jan (advisor) ; Maršík, Ladislav (referee)
Melody extraction is arguably one of the most important and challenging problems in Music Information Retrieval. It is melody that we are likely to recall after listening to a song and so it is one of the most relevant aspects of music. However the presence of accompaniment in songs makes the task hard to address using rule-based methods. During the last years data-driven methods based on deep learning started to outperform methods traditionally used in the field. In this thesis we continue in these efforts and propose three new methods for melody extraction. Among these an architecture called Harmonic Convolutional Neural Network, based on a modification of convolutional neural networks to better capture harmonically related information in an input spectrogram with logarithmic frequency axis, was able to achieve state-of-the-art performance on several publicly available melody datasets. 1
|
guest ::
