Duboko učenje u autonomnim vozilima

master's thesis

• Title (english): Deep Learning for Autonomous Vehicles
• Author: Valentina Fatorić
• Mentor: Vladimir Čeperić (mentor)
• Committee member: Vladimir Čeperić (predsjednik povjerenstva)
• Committee member: Marin Golub (član povjerenstva)
• Committee member: Mirko Poljak (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2020-07-16, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Za autonomnu vožnju potrebno je da vozilo preko kamera i senzora može prepoznati okolinu oko sebe na temelju čega će donijeti odluke. Za to se koristi semantička segmentacija kojom se svakom pikselu slike dodjeljuje određena klasa. Semantička segmentacija je jedan od najtežih problema računalnog vida za kojeg se koristi duboko učenje, odnosno konvolucijske neuronske mreže. U radu su opisana dva trenutno najbolja modela za taj problem i njihovi rezultati dobiveni nad skupom podataka Cityscapes proširenim sa ručno označenim slikama snimljenih po gradu Zagrebu.

Abstract (english)

For autonomous driving, it is necessary that the vehicle can recognize the environment around it through cameras and sensors, on the basis of which it will make decisions. For that it's used semantic segmentation which assigns a certain class to each pixel of the image. Semantic segmentation is one of the most difficult problems of computer vision, for which is used deep learning, ie convolutional neural networks. This paper describes the two currently best models for this problem and their results obtained over the dataset Cityscapes extended with manually annotated images taken around the city of Zagreb.

• Keywords: Duboko učenje; autonomna vozila; semantička segmentacija; konvolucijske mreže; Cityscapes
• Keywords (english): Deep learning; autonomous vehicles; semantic segmentation; convolutional networks; Cityscapes
• Language: croatian
• URN:NBN: urn:nbn:hr:168:907193
• 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: 2023-01-13 12:12:46