prikaz prve stranice dokumenta Efficient facial expression recognition using decision trees and neural networks
  Download
PDF 24.79 MB
doctoral thesis
Efficient facial expression recognition using decision trees and neural networks
Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing, 2021. urn:nbn:hr:168:882946
Gogić, Ivan
University of Zagreb
Faculty of Electrical Engineering and Computing
Department of Telecommunications

Institutional repository: FER Repository

Cite this document

Gogić, I. (2021). Efficient facial expression recognition using decision trees and neural networks (Doctoral thesis). Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing. Retrieved from https://urn.nsk.hr/urn:nbn:hr:168:882946

Gogić, Ivan. "Efficient facial expression recognition using decision trees and neural networks." Doctoral thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2021. https://urn.nsk.hr/urn:nbn:hr:168:882946

Gogić, Ivan. "Efficient facial expression recognition using decision trees and neural networks." Doctoral thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2021. https://urn.nsk.hr/urn:nbn:hr:168:882946

Gogić, I. (2021). 'Efficient facial expression recognition using decision trees and neural networks', Doctoral thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, accessed 27 June 2024, https://urn.nsk.hr/urn:nbn:hr:168:882946

Gogić I. Efficient facial expression recognition using decision trees and neural networks [Doctoral thesis]. Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing; 2021 [cited 2024 June 27] Available at: https://urn.nsk.hr/urn:nbn:hr:168:882946

I. Gogić, "Efficient facial expression recognition using decision trees and neural networks", Doctoral thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, 2021. Available at: https://urn.nsk.hr/urn:nbn:hr:168:882946

Metadata
TitleEfficient facial expression recognition using decision trees and neural networks
Title (croatian)Učinkovito raspoznavanje izraza lica primjenom stabala odlučivanja i neuronskih mreža
AuthorIvan Gogić
MentorIgor Sunday Pandžić (mentor)
MentorJörgen Ahlberg https://orcid.org/0000-0002-6763-5487 (komentor)
Committee memberIgor Sunday Pandžić (član povjerenstva)
GranterUniversity of Zagreb
Faculty of Electrical Engineering and Computing
(Department of Telecommunications)
Zagreb
Defense date and country2021, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Universal decimal classification
(UDC)		004 - Computer science and technology. Computing. Data processing
Abstract
This thesis investigates a facial expression recognition system that estimates the emotional state of subjects from facial images. Such systems demand accurate and fast algorithms that can run in real-time on platforms with limited computational resources. The proposed algorithms bridge the gap between precise but slow methods and fast but less precise methods, combining decision trees and neural networks. The gentle boost decision trees are trained to extract highly discriminative feature vectors for each facial expression around distinct facial landmark points. These sparse binary features are concatenated to jointly optimize facial expression predictions with a shallow neural network architecture. The joint optimization improves the recognition rates of difficult expressions such as fear and sadness. Since the algorithm depends on accurate landmark locations, a novel face alignment method is introduced using gradient boost decision trees and neural networks organized in a cascaded regression framework. The cascade is initialized by a lightweight convolutional neural network to increase robustness while preserving high efficiency. The thesis begins with an introduction to the problem and the motivation for solving it, followed by an explanation of the theoretical background and a systematic overview of related, previous work. Next, novel algorithms for face alignment and facial expression recognition are described and evaluated on relevant public data sets. The results demonstrate high efficiency and competitive accuracy compared to the state-of-the-art methods suitable for power-efficient applications. The final chapter provides concluding remarks of the thesis.
Abstract (croatian)
Ovaj rad istražuje sustav raspoznavanja izraza lica koji predviđa emotivno stanje subjekata iz slika lica. Takvi sustavi zahtijevaju točne i brze algoritme koji se mogu izvršavati u stvarnom vremenu na platformama sa ograničenim računalnim sredstvima. Predloženi algoritmi premošćuju raskorak između preciznih ali sporih metoda i brzih ali manje preciznih metoda kombinacijom stabala odlučivanja i neuronskih mreža. Stabla odlučivanja učena su pomoću gentle boost metode kako bi izlučila visoko diskriminativni vektor značajki za svaki izraz lica u okolini ključnih točaka lica. Dobiveni prorijeđeni binarni vektori spajaju se kako bi se zajednički optimizirala predviđanja izraza lica pomoću plitke arhitekture neuronske mreže. Zajednička optimizacija poboljšava točnost raspoznavanja za teške izraze kao što su strah i ljutnja. Budući da algoritam ovisi o točnim položajima ključnih točaka lica, predložena je nova metoda detekcije ključnih točaka koja koristi gradient boost stabla odlučivanja i neuronske mreže u okviru kaskadne regresije. Kaskada se inicijalizira pomoću optimizirane konvolucijske neuronske mreže kako bi se poboljšala robusnost metode, a zadržala visoka razina učinkovitosti. Doktorski rad počinje uvodom u problem i motivaciju za njegovo rješavanje, a slijedi ga obrazloženje teoretske pozadine te sistematski pregled povezanog prethodnog istraživanja. Nakon toga, opisani su novi algoritmi za detekciju ključnih točaka i raspoznavanje izraza lica te testirani na relevantnim javnim skupovima podataka. Dobiveni rezultati pokazuju visoku učinkovitost i konkurentnu točnost u usporedbi sa najboljim metodama te prikladnost metode za energetski učinkovite primjene. Posljednje poglavlje zaključuje doktorski rad.
Keywords
Keywords (croatian)
Languageenglish
URN:NBNurn:nbn:hr:168:882946
Study programmeTitle: Computer Science
Study programme type: university
Study level: postgraduate
Academic / professional title: Doktor znanosti (Doktor znanosti)
Catalog URLhttp://lib.fer.hr/cgi-bin/koha/opac-detail.pl?biblionumber=52611
Type of resourceText
Extentxiii, 106 str.
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2022-04-27 11:43:39