Thesis Details
Hluboké neuronové sítě pro klasifikaci objektů v obraze
This paper deals with classifying objects using deep neural networks. Whole scene segmentation was used as main algorithm for the classification purpose which works with video sequences and obtains information between two video frames. Optical flow was used for getting information from the video frames, based on which features maps of a~neural network are warped. Two neural network architectures were adjusted to work with videos and experimented with. Results of the experiments show, that using videos for image segmentation improves accuracy (IoU) compared to the same architecture working with images.
image segmentation, monocular camera, deep neural networks, video, optical flow, warping, Cityscapes, Keras, Tensorflow
Hradiš Michal, Ing., Ph.D. (DCGM FIT BUT), člen
Rogalewicz Adam, doc. Mgr., Ph.D. (DITS FIT BUT), člen
Sochor Jiří, prof. Ing., CSc. (FI MUNI), člen
Španěl Michal, doc. Ing., Ph.D. (DCGM FIT BUT), člen
Zemčík Pavel, prof. Dr. Ing. (DCGM FIT BUT), člen
@mastersthesis{FITMT19405,
author = "Tom\'{a}\v{s} Mlynari\v{c}",
type = "Master's thesis",
title = "Hlubok\'{e} neuronov\'{e} s\'{i}t\v{e} pro klasifikaci objekt\r{u} v obraze",
school = "Brno University of Technology, Faculty of Information Technology",
year = 2018,
location = "Brno, CZ",
language = "czech",
url = "https://www.fit.vut.cz/study/thesis/19405/"
}