Klasifikacija slika korištenjem transformatorskih mreža

undergraduate thesis

• Title (english): Image Classification Using Transformer Networks
• Author: Matko Barbić
• Mentor: Sven Lončarić (mentor)
• Committee member: Sven Lončarić (predsjednik povjerenstva)
• Committee member: Luka Posilović (član povjerenstva)
• Committee member: Donik Vršnak (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-09-13, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Kompjuterska vizija je važno područje umjetne inteligencije. Od 2012. konvolucijske neuronske mreže su najčešće korišteni oblik umjetne inteligencije za rješavanje problema u području računalnog vida. U 2021. transformer arhitektura, koja se do tada uglavnom koristila za probleme procesiranja prirodnog jezika, je adaptirana za rješavanje problema računalnog vida. U ovom radu smo proučili mehanizam pažnje koji se koristi u transformerskim arhitekturama, implementirali dvije najpopularnije transformerske arhitekture za rad na slikama i usporedili rezultate s modernom konvolucijskom neuronskom mrežom.

Abstract (english)

Computer vision is an important area of artificial intelligence. Since 2012. convolutional neural networks are most common form of solving problems in the computer vision field. In 2021. transformer architecture, which was until then mostly used for natural language processing, was adapted for solving computer vision tasks. In this paper we investigated the attention mechanism used in transformer architectures, implemented two of the most popular transformer architectures used on images and compared the results with a modern convolutional neural network.

• Keywords: pažnja; računalni vid; raspoznavanje slika; transformer
• Keywords (english): attention; computer vision; image classification; transformer
• Language: croatian
• URN:NBN: urn:nbn:hr:168:547052
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-15 09:44:40