An Evaluation Method of Speech Communication Jamming Effect in Low SNR Environment

The commonly used objective evaluation method of speech communication jamming effect based on characteristic parameters requires the time alignment between the original speech and disturbed speech. In the environment of low signal-to-noise ratio, due to the influence of jamming signal, the time alignment requirements are often unable to be met, resulting in the reduction of Pearson coefficient of objective evaluation. This method applies the automatic speech recognition technology based on deep learning to the objective evaluation of speech communication jamming effect. Through the speech recognition of the original speech and the disturbed speech, the word correct rate is established. The mapping relationship between the word correct rate and the subjective evaluation is established, and the word correct rate is converted into the subjective evaluation value, so as to determine the jamming effect. This method avoids the requirement of time alignment between original speech and disturbed speech, and improves the Pearson coefficient of objective evaluation.