Objective Evaluation of Speech Communication Jamming Effect based on Composite Evaluation

Abstract

In the objective evaluation of speech communication interference effect, a single objective evaluation method has the problem of poor scene adaptability. In some communication scenarios, the correlation is good, while in other communication scenarios, the correlation is poor. To solve this problem, this paper uses multivariate adaptive regression splines method to composite PESQ, MFCC and other objective evaluation methods. Based on complementary design, the objective evaluation methods of improving the scene adaptability is achieved. The experimental data shows that the MARS-based composition evaluation method has better performance than a single objective evaluation method.