Semantička segmentacija osoba dubokim konvolucijskim modelom

master's thesis

• Title (english): Semantic Segmentation of Humans Using a Deep Convolutional Model
• Author: Natalija Šegota
• Mentor: Marin Šilić (mentor)
• Committee member: Marin Šilić (predsjednik povjerenstva)
• Committee member: Goran Delač (član povjerenstva)
• Committee member: Klemo Vladimir (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-07-04, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj rad bavi se problemom klasifikacije svakog piksela slike. Objašnjeni su modeli zasnovani na dubokim neuronskim mrežama, koje se koristi pri semantičkoj segmentaciji slike. Sustav za semantičku segmentaciju osoba programski je ostvaren, a zatim i procijenjen prikladnim metodama. Procjena je provedena na javno dostupnom skupu slika računanjem omjera presjeka i unije oznake slike i predikcije dobivene modelom.

Abstract (english)

This paper examines the problem of image pixel by pixel classification. The models for semantic image segmentation based on deep neural networks have been explained as well. The semantic segmentation of human system has been implemented and evaluated by appropriate methods. The evaluation was made on a publicly available dataset of images using by calculating the ratio of the overlap area to the combined area of prediction and ground truth.

• Keywords: semantička segmentacija; duboko učenje; duboke neuronske mreže; konvolucijske neuronske mreže; klasifikacija; lokalizacija; mape značajki
• Keywords (english): semantic segmentation; deep learning; deep neural networks; convolution neural networks; classification; localization; feature maps
• Language: croatian
• URN:NBN: urn:nbn:hr:168:579376
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-10-11 10:38:28