Please use this identifier to cite or link to this item:https://hdl.handle.net/20.500.12259/43348
Type of publication: Tezės kituose recenzuojamuose leidiniuose / Theses in other peer-reviewed publications (T1e)
Field of Science: Informatika / Computer science (N009)
Author(s): Šiugždaitė, Roma;Norvaišas, Saulius;Saudargienė, Aušra;Grecucci, Alessandro
Title: How natural system anticipates emotion recognition
Is part of: CHAOS 2011 : 4th chaotic modeling and simulation international conference, May 31-June 3, 2011, Agios Nikolaos, Crete, Greece : book of abstracts. Agios Nikolaos, Crete, Greece, 2011
Extent: p. 133-133
Date: 2011
Keywords: Facial emotion recognition;FER;Mental model;Dynamical system
Abstract: In this research paper we hypothesize the human facial emotions recognition (FER) process is related to gaze “jumping” and to emotional level of observer itself. Evolutionary speaking it’s vitally important to recognize emotions quickly and precisely in order to evaluate the situation and react to it adequately. We suppose that facial emotion recognition is a process of convergence of induced neural zones into one of many local attractors, which represent mental models or behavioral schemes as a reaction to certain emotions. The convergence speed is growing if visual FER part (“data collection”) is leaded by gaze fluctuations between emotion’ representing segments (for example, facial parts). The higher uncertainty of visage is, the bigger gaze fluctuations there are. In the case of factitious emotional expressions, FER is more difficult if observer is better trained and gaze “jumping” is minimal. Gaze “jumping” in nonstandard situations of untrained person (like a child) grows independently if it leads or not to only attractor (i.e. the emotion is recognized or is not recognized). Consequently the gaze “drifting” is related to experience and training, but an increased gaze fluctuations means transition from one local attractor to another improving FER process. Therefore learned emotions serves to recognize better the other emotions, by increased gaze „drifting“ - emotions participate as catalyst of recognition process. Also asymmetry evokes gaze fluctuations and even a small one anticipates recognition process. During the experiment (unnatural facial emotions recognition) with children, gaze “jumping” is higher even if any decision follows from that. Meanwhile gaze oscillations of adults are smaller and lead to certain identification cluster (even misguided)
Internet: http://www.cmsim.org/images/CHAOS2011-Book_of_Abstracts.pdf
http://www.cmsim.org/images/CHAOS2011-Book_of_Abstracts.pdf
Affiliation(s): Informatikos fakultetas
Taikomosios informatikos katedra
Vytauto Didžiojo universitetas
Appears in Collections:Universiteto mokslo publikacijos / University Research Publications

Files in This Item:
marc.xml7.64 kBXMLView/Open

MARC21 XML metadata

Show full item record

Page view(s)

132
checked on Dec 9, 2019

Download(s)

10
checked on Dec 9, 2019

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.