magistrsko delo
Patrik Kokol (Author), Aleš Holobar (Mentor), Boris Cigale (Co-mentor)

Abstract

Cilj magistrskega dela je bil implementirati sistem za avtomatsko nastavljanje višine mize uporabniku z uporabo globinskega senzorja. S pomočjo konvolucijskih nevronskih mrež smo uporabnika detektirali znotraj globinske scene ter tako izračunali njegovo višino. Na podlagi izračunane višine smo nastavili tudi pravilno višino mize za stoječi ali sedeči položaj uporabnika. Pri detekciji osebe v globinski sceni smo dosegli 85,5% natančnost po metriki preseka proti uniji (angl. Intersection over Union - IoU) in 92,3% natančnost po metriki F1. Napaka pri izračunu višine je v povprečju znašala 0,092 metra.

Keywords

nevronske mreže;konvolucijske nevronske mreže;umetna inteligenca;globoko učenje;detekcija osebe;magistrske naloge;

Data

Language: Slovenian
Year of publishing:
Typology: 2.09 - Master's Thesis
Organization: UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher: [P. Kokol]
UDC: 004.85:004.932(043.2)
COBISS: 27324931 Link will open in a new window
Views: 496
Downloads: 45
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Detection of single person in depth image using convolutional neural networks
Secondary abstract: The goal of this thesis was the implementation of a system for automatic table height adjustment using a depth sensor. With the help of convolutional neural networks we have detected the user in the depth scene captured by the depth sensor. Using the detection we have also calculated the height of the user and based on his height we have adjusted the table height accordingly to the needs of the user (standing or sitting). We measured the user detection based on two metrics intersection over union (IoU) and F1 score. We achieved IoU value of 85,5% and F1 score of 92,3%. The average height calculation error was 0,092 meters.
Secondary keywords: neural networks;convolutional neural network;artificial inteligence;deep learning;human detection;
Type (COBISS): Master's thesis/paper
Thesis comment: Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Računalništvo in informacijske tehnologije
Pages: XI, 49 str.
ID: 11606947