The management of mental health in a smart medical dialogue system based on a two-stage attention speech enhancement module

Abstract

In response to the current problems in artificial intelligence medical conversations, this project intends to study a new speech reinforcement algorithm based on two-level attention mechanism to improve its correct understanding and execution ability in verbal communication. Firstly, the paper conducted theoretical and research on commonly used speech enhancement algorithms, and studied them; On this basis, a speech enhancement algorithm based on a two-level attention reinforcement network is studied, and the algorithm is applied to intelligent medical conversations. The experiment shows that the signal-to-noise ratio obtained by using a two-level attention enhancement algorithm is higher than that of the attention enhancement algorithm for a 3 dB channel, with a 3 % improvement in signal-to-noise ratio, which is the smallest among the four algorithms. In addition, 98.95 % of students reported participating in 13 or more tutoring classes, accounting for 54 %. Based on the above research, the proposed speech enhancement method in this article is feasible and effective.