Detekcija i raspoznavanje lica u slikama

undergraduate thesis

• Title (english): Face detection and recognition in images
• Author: Marko Bagarić
• Mentor: Zoran Kalafatić (mentor)
• Committee member: Zoran Kalafatić (predsjednik povjerenstva)
• Committee member: Zorana Doždor (član povjerenstva)
• Committee member: Igor Stančin (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-15, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Prepoznavanje lica trenutno je u rastu popularnosti u primjeni u području osiguranja, posebice u telefonima. Prisutne su mnoge tehnike prepoznavanja lica, a zadaća projekta je implementacija jednog od načina, opis prepoznavanja i usporedba s ostalim pristupima problematici.
OpenCV algoritam s LBPH deskriptorima, iako ne pretjerano velike točnosti, daje jednostavan uvid u problematiku implementacije prepoznavanja lica i raznih open-source rješenja koja su primjenjiva. U samom projektu sigurnost je veoma varirala zbog male količine prikupljenih podataka te nemogućnosti korištenja dodatnih metoda podjele baze podataka za testiranje.
Rezultati daju FaceNetu najbolju ocjenu s obzirom na predikciju, ali ne i za vrijeme potrebno da se predikcija ostvari. Trenutno baza podataka FaceNeta i napredni algoritam za prepoznavanje omogućavaju konkurentnost naprema ostalim algoritmima.

Abstract (english)

Face recognition is growing in popularity in the security of applications, especially in telephones. Many facial recognition techniques are present, and the task of the project is to implement one of the methods, describe the recognition and comparation of different approaches to problem.
OpenCV algorithm s LBPH descriptors, although not overly accurate, provides a simple insight into the implementation of face recognition and various open-source solutions that are applicable. In the project itself, confidence varied greatly due to the small amount of data collected and the inability to use additional database splitting methods for testing.
The results give FaceNet the best rating in terms of prediction, but not in the time it takes for the prediction to materialize. Currently, the FaceNet database and advanced recognition algorithm allow further competition against other algorithms.

• Keywords: Strojno učenje; Prepoznavanje lica; Umjetne neuronske mreže
• Keywords (english): Machine learning; Face recognition; Artificial neuron network
• Language: croatian
• URN:NBN: urn:nbn:hr:168:363747
• 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 09:42:40