Application Research of Intelligent Educational Robot in Interactive Teaching of Secondary Vocational Class

Aiming at the problem of human-computer interaction of teaching robots in secondary vocational schools and aiming to improve the teaching quality, this paper conducts a study on the speech enhancement of secondary vocational school students under the two-level attention mechanism. This paper first conducts theoretical and experimental research on the currently commonly used speech enhancement algorithms. Then, a two-level attention enhancement neural network is established, and the two-level attention enhancement model is added to the conversation of the secondary vocational teaching robot for experimental verification. The experiment proves that compared with the high-resolution channel attention enhancement network, the signal-to-noise comparator (SSNR) of the improved algorithm proposed in this paper can reach 3.15 dB, and the loss value is reduced to 0.985. At the same time, after the noise processing, its time domain waveform is clearer. In short, the speech enhancement algorithm studied in this paper is feasible and effective, which can enable the secondary vocational teaching robot to understand and complete text commands more accurately.