A New Neural Network Measure for Objective Speech Quality Evaluation

A new measure for objective speech quality evaluation based on the improved generalized congruence neural network (GCNN/OSQE) is proposed, which needs less training time and has better performance. Compared with radial basis function neural network for objective speech quality evaluation measure (RBFNN/OSQE), besides owning all the merits of RBFNN/OSQE, GCNN/OSQE has many more merits: higher correlation, smaller standard deviation, and saving about 1/3 training time. In all, the results of speech quality assessment show that the proposed GCNN/OSQE is feasible and effective.