Scientific and Research Laboratory "Visual Information Systems"
The members of the Laboratory have teaching and research activities in the area of: Digital Signal Processing, Image and Video Processing, Image and Sound Compression, Pattern recognition and Neural Networks, Multimedia Systems, Still and Video Image Face Analysis and Recognition, Multimodal Biometrics. Head of the Laboratory is Prof. Ognian Boumbarov, PhD.
In the last 10 years the work of the Research Laboratory "Visual Information Systems" is connected to the development of methods, algorithms and processing systems, coding and analysis of images and sound research in multimodal biometric systems and systems using human-computer interface. The staff of the S & R lab has worked on 5 projects related with the topic of Face Emotion Recognition. Collective efforts are focused primarily on the development of methods and algorithms for 2D and 3D Face detection, extraction of feature vectors of individual faces in the subspaces (PCA, ICA, LDA, spectral regression SR) and application of combined classifiers in the recognition stage.
Combined structures of neural network classifiers are developed based on radial basis function neural networks, convolutional neural networks and support vector machines. The phase of segmentation of human faces is determined by using a combined method using the color of the skin, modeled with 2D and 3D histograms and Gaussian mixture models. The geometric information about the shape and the main facial components - mouth, eyes and nose it is done with the verification using convolutional neural network. The team has created an algorithm for managing screen menus with gaze tracking. On the topic Face Emotion Recognition in the past five years has been published more than 40 articles in prestigious journals and scientific conferences.