Procjena poze sportaša temeljeno na dubokom učenju

undergraduate thesis

• Title (english): Athletes Pose Estimation Based on Deep Learning
• Author: Jure Spajić
• Mentor: Vladimir Čeperić (mentor)
• Committee member: Vladimir Čeperić (predsjednik povjerenstva)
• Committee member: Josip Šarić (član povjerenstva)
• Committee member: Marin Kačan (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-04, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Tema rada je procjena poze sportaša temeljeno na metodama dubokog učenja. Duboko učenje je podgrana strojnog učenja koja sustav obrađuje na način koji je sličan ljudskom mozgu. Kao prikladan model odabrane su konvolucijske neuronske mreže, koje
su se pokazale izuzetno učinkovitima u analizi i klasifikaciji slika. Opisan je skup podataka Leeds Sport Pose koji je podijeljen na skup za učenje, validaciju i testiranje, kako
bi se osiguralo pravilno treniranje modela i procjena njegove izvedbe. Model je treniran da prepoznaje pravilne i nepravilne stavove kod surfera. Na kraju su dani prijedlozi
poboljšanja i nadogradnje modela.

Abstract (english)

The topic of this project is the assessment of athlete poses based on deep learning
methods. Deep learning is a subfield of machine learning that processes data in a way
similar to the human brain. Convolutional neural networks (CNNs) have been selected
as the appropriate model, as they have proven to be extremely effective in image analysis and classification. The Leeds Sport Pose dataset is described, which is divided into
training, validation, and testing sets to ensure proper model training and performance
evaluation. The model was trained to recognize correct and incorrect stances in surfers.
Finally, suggestions for model improvement and upgrades are provided.

• Keywords: duboko učenje; konvolucijske neuronske mreže; procjena poze; tehnike učenja; Leeds Sport Pose; TensorFlow; ključne točke
• Keywords (english): deep learning; convolutional neural networks; pose estimation; learning techniques; Leeds Sport Pose; TensorFlow; key points
• Language: croatian
• URN:NBN: urn:nbn:hr:168:438963
• 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: 2024-10-12 22:08:37