To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Face emotion recognition system based on fuzzy logic using algorithm improved Particle Swarm |
|
Author |
Seyed mostafa sharifi, Marjan abdeyazdan |
| Citation |
Vol. 16 No. 7 pp. 157-166 |
|
Abstract |
Recognition of facial emotions has important role in HCI. One of the components of the system for facial emotion Recognition is classification. For classification with the fuzzy logic system, setting parameters of membership functions has significant importance. In this research, first for extraction of feature, two extractions of feature linear subspace projection and nonlinear subspace projection toolboxes named pretty helpful development toolboxes were used. For classification of the resultant data from feature extraction, the fuzzy inference system was used. For setting the parameters of membership functions, the improved PSO algorithm was taken advantage of to be able to increase the precision of classification. Precision of system classification was calculated with each of the PSO, ACOR, Genetic and SA algorithms. Next, the precision of classification resulting from combination of PSO with each of these algorithms was calculated. Results show that combination of PSO with Genetic algorithm yields a more advantageous result compared to the other algorithms. In this research, also the precision of recognition of the emotion of contemptuousness was calculated with the proposed system which up to now has not been evaluated in any article on set of emotions. Experimental results report a mean precision of measurement of the system for Recognition of seven emotions to be 98.32 percent. Additionally, using the proposed system, the rate of Recognition of the feeling of contemptuousness was reported at 99.25 percent, where in reports obtained from fuzzy deduction system, classification with Genetic algorithm, has reported classification precisions of 100 percent as well. |
|
Keywords |
Emotion face recognition system, Fuzzy inference system, Particle swarm optimization (PSO), Ant colony optimization real (ACOR), Genetic Algorithm (GA), Simulated annealing (SA). |
|
URL |
