Intelligent Recognition of Aircraft Type Using Audio Feature
Ruida Ye, Weijie Wang, Ziyang Wang, Jie Wen Zhao · 2021
Aiming at the shortcomings of existing aircraft type recognition methods, an intelligent type recognition method based on audio features is proposed. The improved self-attention mechanism algorithm is used to intelligently recognize and classify aircraft. In the existing self-attention mechanism, three weight matrices are obtained through linear transformation, which is not conducive to the model learning complex audio features. In the existing self-attention mechanism, non-linear factors are introduced into the network model and a non-linear self-attention mechanism algorithm is proposed. Using the collected 9 types of aircraft audio data, after a series of aircraft audio features, it is used as the data set of the algorithm, and the recognition ability of the algorithm is verified through experiments.