Araştırma Çıktıları | WoS | Scopus | TR-Dizin | PubMed
Permanent URI for this communityhttps://hdl.handle.net/20.500.14719/1741
Browse
6 results
Search Results
Publication Metadata only INTERSPEECH 2009 emotion recognition challenge evaluation, INTERSPEECH 2009 duygu tanima yarişmasi deǧerlendirmesi(2010) Bozkurt, Elif; Erzin, Engin; Erdem, Cigdem Eroglu; Erdem, Tanju Tanju; Bozkurt, Elif, Koç Üniversitesi, Istanbul, Turkey; Erzin, Engin, Koç Üniversitesi, Istanbul, Turkey; Erdem, Cigdem Eroglu, Bahçeşehir Üniversitesi, Istanbul, Turkey; Erdem, Tanju Tanju, Özyeğin Üniversitesi, Istanbul, TurkeyIn this paper we evaluate INTERSPEECH 2009 Emotion Recognition Challenge results. The challenge presents the problem of accurate classification of natural and emotionally rich FAU Aibo recordings into five and two emotion classes. We evaluate prosody related, spectral and HMM-based features with Gaussian mixture model (GMM) classifiers to attack this problem. Spectral features consist of mel-scale cepstral coefficients (MFCC), line spectral frequency (LSF) features and their derivatives, whereas prosody-related features consist of pitch, first derivative of pitch and intensity. We employ unsupervised training of HMM structures with prosody related temporal features to define HMM-based features. We also investigate data fusion of different features and decision fusion of different classifiers to improve emotion recognition results. Our two-stage decision fusion method achieves 41.59 % and 67.90 % recall rate for the five and two-class problems, respectively and takes second and fourth place among the overall challenge results. ©2010 IEEE. © 2011 Elsevier B.V., All rights reserved.Publication Metadata only A comparison of geometrical facial features for affect recognition, Duygu tanima i̇çi̇n geometri̇k yüz özni̇teli̇kleri̇ni̇n karşilaştirilmasi(2011) Ulukaya, Sezer; Erdem, Cigdem Eroglu; Ulukaya, Sezer, Bahçeşehir Üniversitesi, Istanbul, Turkey; Erdem, Cigdem Eroglu, Bahçeşehir Üniversitesi, Istanbul, TurkeyIn this work, we compare two different geometric feature extraction methods derived from coordinates of facial points tracked by Active Appearance Models. The compared feature extraction methods differ in their use of coordinates or distances between facial points and whether they use the information of a neutral facial expression. Experiments on the extended Cohn-Kanade database show that the coordinate-based features using the neutral frame information gives the best emotion recognition results (%94) using a SVC classifier with a polynomial kernel. © 2011 IEEE. © 2011 Elsevier B.V., All rights reserved.Publication Metadata only RANSAC-based training data selection on spectral features for emotion recognition from spontaneous speech(2011) Bozkurt, Elif; Erzin, Engin; Erdem, Cigdem Eroglu; Erdem, Tanju Tanju; Bozkurt, Elif, College of Engineering, Koç Üniversitesi, Istanbul, Turkey; Erzin, Engin, College of Engineering, Koç Üniversitesi, Istanbul, Turkey; Erdem, Cigdem Eroglu, Department of Electrical and Electronic Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Erdem, Tanju Tanju, Department of Electrical and Electronic Engineering, Özyeğin Üniversitesi, Istanbul, TurkeyTraining datasets containing spontaneous emotional speech are often imperfect due the ambiguities and difficulties of labeling such data by human observers. In this paper, we present a Random Sampling Consensus (RANSAC) based training approach for the problem of emotion recognition from spontaneous speech recordings. Our motivation is to insert a data cleaning process to the training phase of the Hidden Markov Models (HMMs) for the purpose of removing some suspicious instances of labels that may exist in the training dataset. Our experiments using HMMs with Mel Frequency Cepstral Coefficients (MFCC) and Line Spectral Frequency (LSF) features indicate that utilization of RANSAC in the training phase provides an improvement in the unweighted recall rates on the test set. Experimental studies performed over the FAU Aibo Emotion Corpus demonstrate that decision fusion configurations with LSF and MFCC based classifiers provide further significant performance improvements. © 2011 Springer-Verlag. © 2011 Elsevier B.V., All rights reserved.Publication Metadata only Feature extraction for facial expression recognition by canonical correlation analysis, Kanoni̇k korelasyon anali̇zi̇ i̇le yüz i̇fadesi̇nden duygu tanima i̇çi̇n özni̇teli̇ k çikarimi(2012) Sakar, C. Okan; Kursun, Olcay; Karaali, Ali; Erdem, Cigdem Eroglu; Sakar, C. Okan, Bahçeşehir Üniversitesi, Istanbul, Turkey; Kursun, Olcay, Istanbul Üniversitesi, Istanbul, Turkey; Karaali, Ali, Bahçeşehir Üniversitesi, Istanbul, Turkey; Erdem, Cigdem Eroglu, Bahçeşehir Üniversitesi, Istanbul, TurkeyAlthough several methods have been proposed for fusing different image representations obtained by different preprocessing methods for emotion recognition from the facial expression in a given image, the dependencies and relations among them have not been much investigated. In this study, it has been shown that covariates obtained by Canonical Correlation Analysis (CCA) that extracts relations between different representations have high predictive power for emotion recognition. As high prediction accuracy can be achieved using a small number of features extracted by it, CCA is considered to be a good dimensionality reduction method. For our simulations, we used the CK+ database and showed that covariates obtained from difference-images and geometric-features representations have high prediction accuracy. © 2012 IEEE. © 2012 Elsevier B.V., All rights reserved.Publication Metadata only Estimation of the neutral face shape using Gaussian mixture models(2012) Ulukaya, Sezer; Erdem, Cigdem Eroglu; Ulukaya, Sezer, Department of Electrical and Electronic Engineering, Boğaziçi Üniversitesi, Bebek, Turkey, Department of Electrical and Electronic Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Erdem, Cigdem Eroglu, Department of Electrical and Electronic Engineering, Bahçeşehir Üniversitesi, Istanbul, TurkeyWe present a Gaussian Mixture Model (GMM) fitting method for estimating the unknown neutral face shape for frontal facial expression recognition using geometrical features. Subtracting the estimated neutral face, which is related to the identity-specific component of the shape leaves us with the component related to the variations resulting from facial expressions. Experimental results on the Extended Cohn-Kanade (CK+) database show that subtracting the estimated neutral face shape gives better emotion recognition rates as compared to classifying the geometrical facial features directly, when the person-specific neutral face shape is not available. We also experimentally evaluate two different geometric facial feature extraction methods for emotion recognition. The average emotion recognition rates achieved with the proposed neutral shape estimation method and coordinate based features is 88%, which is higher than the baseline results presented in the literature, although we do not use the person-specific neutral shapes (94% if we use), and any appearance based features. © 2012 IEEE. © 2012 Elsevier B.V., All rights reserved.Publication Metadata only A method for extraction of affective audio-visual facial clips from movies, Filmlerden duygusal yüz ifadeleri içeren video klipleri elde etmek için bir yöntem(2013) Turan, Çigdem; Kansin, Can; Zhalehpour, Sara; Aydin, Zafer; Erdem, Cigdem Eroglu; Turan, Çigdem, Bahçeşehir Üniversitesi, Istanbul, Turkey; Kansin, Can, Bahçeşehir Üniversitesi, Istanbul, Turkey; Zhalehpour, Sara, Bahçeşehir Üniversitesi, Istanbul, Turkey; Aydin, Zafer, Bahçeşehir Üniversitesi, Istanbul, Turkey; Erdem, Cigdem Eroglu, Bahçeşehir Üniversitesi, Istanbul, TurkeyIn order to design algorithms for affect recognition from facial expressions and speech, audio-visual databases are needed. The affective databases used by researchers today are generally recorded in laboratory environments and contain acted expressions. In this work, we present a method for extraction of audio-visual facial clips from movies. The database collected using the proposed method contains English and Turkish clips and can easily be extended for other languages. We also provide facial expresssion recognition results, which utilize local phase quantization based feature extraction and a support vector machine. Due to larger number of features compared to the number of examples, the affect recognition accuracy improves significantly when feature selection is also performed. © 2013 IEEE. © 2013 Elsevier B.V., All rights reserved.