Qualification: 
Ph.D, M.Tech
p_suja@blr.amrita.edu

Dr. Suja P. currently serves as Assistant Professor at the Department of Computer Science, Amrita School of Engineering. She successfully defended her PhD in "Robust Emotion Recognition Techniques from Facial Expressions Using Images & Videos" under the guidance of Dr. Shikha Tripathi. Her areas of research include Image Processing, Computer Graphics, Emotion Recognition.

Publications

Publication Type: Journal Article

Year of Publication Publication Type Title

2017

Journal Article

SaSai Prathusha, Dr. Suja P., Dr. Shikha Tripathi, and Louis, Rc, “Emotion recognition from facial expressions of 4D videos using curves and surface normals”, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10127 LNCS, pp. 51-64, 2017.[Abstract]


In this paper, we propose and compare three methods for recognizing emotions from facial expressions using 4D videos. In the first two methods, the 3D faces are re-sampled by using curves to extract the feature information. Two different methods are presented to resample the faces in an intelligent way using parallel curves and radial curves. The movement of the face is measured through these curves using two frames: neutral and peak frame. The deformation matrix is formed by computing the distance point to point on the corresponding curves of the neutral frame and peak frame. This matrix is used to create the feature vector that will be used for classification using Support Vector Machine (SVM). The third method proposed is to extract the feature information from the face by using surface normals. At every point on the frame, surface normals are extracted. The deformation matrix is formed by computing the Euclidean distances between the corresponding normals at a point on neutral and peak frames. This matrix is used to create the feature vector that will be used for classification of emotions using SVM. The proposed methods are analyzed and they showed improvement over existing literature. © Springer International Publishing AG 2017.

More »»

2017

Journal Article

Dr. Suja P. and Dr. Shikha Tripathi, “Geometrical approach for emotion recognition from facial expressions using 4D videos and analysis on feature-classifier combination”, International Journal of Intelligent Engineering and Systems, vol. 10, pp. 30-39, 2017.[Abstract]


Emotion recognition from facial expressions using videos is important in human computer communication where the continuous changes in face movements need to be recognized efficiently. In this paper, a method using the geometrical based approach for feature extraction and recognition of six basic emotions has been proposed which is named as GAFCI (Geometrical Approach for Feature Classifier Identification). Various classifiers, Support Vector Machine (SVM), Random Forest, Naïve Bayes and Neural Networks are used for classification, and the performances of all the chosen classifiers are compared. Out of the 83 feature points provided in the BU4DFE database, optimum feature points are identified by experimenting with several sets of feature points. Suitable "feature-classifier" combination has been obtained by varying the number of feature points, classifier parameters, and training and test samples. A detailed analysis on the feature points and classifiers has been performed to learn the relationship between distance parameters and classification of emotions. The results are compared with literature and found to be encouraging.

More »»

2015

Journal Article

Dr. Suja P. and Dr. Shikha Tripathi, “Analysis of emotion recognition from facial expressions using spatial and transform domain methods”, International Journal of Advanced Intelligence Paradigms, vol. 7, pp. 57-73, 2015.[Abstract]


Facial expressions are non-verbal signs that play an important role in interpersonal communications. There are six basic universally accepted emotions viz., happiness, surprise, anger, sadness, fear and disgust. An emotion recognition system is used for recognising different expressions from the facial images/videos and classifying them into one of the six basic emotions. Spatial domain methods are more popularly used in literature in comparison to transform domain methods. In this paper, two approaches viz., cropped face and whole face methods for feature extraction are implemented separately on the images taken from Cohn-Kanade (CK) and JAFFE databases. Classification is performed using K-nearest neighbour and neural network. The results are compared and analysed. The results suggest that transform domain techniques yield better accuracy than spatial domain techniques and cropped face approach outperforms whole face approach for both the databases for few feature extraction methods. Such systems find application in human computer interaction, entertainment industry and could be used for clinical diagnosis. Copyright © 2015 Inderscience Enterprises Ltd.

More »»

Publication Type: Conference Paper

Year of Publication Publication Type Title

2017

Conference Paper

and Dr. Suja P., “Emotion Recognition from 3D Videos using Optical Flow Method”, in International Conference On Smart Technologies For Smart Nation (SmartTechCon2017), Reva University, Bengaluru, 2017.

2017

Conference Paper

S. K.M and Dr. Suja P., “A Geometric Approach for Recognizing Emotions From 3D Images with Pose Variations”, in International Conference On Smart Technologies For Smart Nation (SmartTechCon2017), Reva University, Bengaluru, 2017.

2014

Conference Paper

Dr. Suja P., Tripathi, S., and Deepthy, J., “Emotion Recognition From Facial Expressions Using Frequency Domain Techniques”, in First International Symposium on Signal Processing and Intelligent Recognition Systems - SIRS 2014, IIITMK- Technopark, Trivandrum, India, 2014.[Abstract]


An emotion recognition system from facial expression is used for recognizing expressions from the facial images and classifying them into one of the six basic emotions. Feature extraction and classification are the two main steps in an emotion recognition system. In this paper, two approaches viz., cropped face and whole face methods for feature extraction are implemented separately on the images taken from Cohn-Kanade (CK) and JAFFE database. Transform techniques such as Dual – Tree Complex Wavelet Transform (DT-CWT) and Gabor Wavelet Transform are considered for the formation of feature vectors along with Neural Network (NN) and K-Nearest Neighbor (KNN) as the Classifiers. These methods are combined in different possible combinations with the two aforesaid approaches and the databases to explore their efficiency. The overall average accuracy is 93{%} and 80{%} for NN and KNN respectively. The results are compared with those existing in literature and prove to be more efficient. The results suggest that cropped face approach gives better results compared to whole face approach. DT-CWT outperforms Gabor wavelet technique for both classifiers.

More »»
207
PROGRAMS
OFFERED
5
AMRITA
CAMPUSES
15
CONSTITUENT
SCHOOLS
A
GRADE BY
NAAC, MHRD
8th
RANK(INDIA):
NIRF 2018
150+
INTERNATIONAL
PARTNERS