Title :
Research of expression recognition base on optimized BP neural network
Author_Institution :
Inf. Sci. & Technol., Tianjin Univ. of Finance & Econ., Tianjin, China
Abstract :
This paper combined Genetics algorithm with BP neural network, mixed-training neural networks, optimized the classification process. Genetic algorithm optimized BP network structure and parameters, found better search space in the analytic space. Then BP network searched the optimal solution in the small search space. Finally, two experiment results indicate that optimized BP neural network training is a effective method of study expression classification and recognition.
Keywords :
backpropagation; emotion recognition; genetic algorithms; neural nets; pattern classification; analytic space; backpropagation; classification process optimization; expression recognition; genetics algorithm; mixed training neural networks optimization; optimized BP Neural Network; search space; Algorithm design and analysis; Biological cells; Character recognition; Convergence; Evolution (biology); Face recognition; Genetic algorithms; Genetic mutations; Neural networks; Optimization methods; BP neural network; facial expression recognition; genetics algorithm;
Conference_Titel :
Industrial Engineering and Engineering Management, 2009. IE&EM '09. 16th International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-3671-2
Electronic_ISBN :
978-1-4244-3672-9
DOI :
10.1109/ICIEEM.2009.5344321