Design and Application of Virtual Simulation Teaching System under Artificial Intelligence Orientation

This paper builds a virtual simulation teaching system based on the Unity simulation platform, which contains a network connection module, resource loading module, and so on. Using the combination of natural interaction and virtual simulation teaching system, the design of Kinect depth information and SVM classifier for gesture recognition classification is integrated with the virtual simulation teaching scene to realize cross-platform collaborative interaction. The virtual simulation teaching process involves a large number of gesture interaction processes, using mf-score multi-feature selection of features that play a vital role in the classification of the data set, eliminating the raw noise of the customized gesture data and improving the contrast of the gesture image. The results of the practical application of the virtual simulation teaching system show that the discipline to which the virtual simulation course belongs (0.008) and the time of identification (0.047) have a significant effect on its application activity. The multisensory, immersive, and interactive nature of virtual simulation teaching (P < 0.001) all have a highly significant positive effect on learners’ foreign language useability. The foreign language assessment score at the end of the virtual simulation course is (92.74±3.32), which is significantly different from the traditional teaching method (P<0.01). The teaching system in this paper meets the requirements of most learners in terms of interactive interface, function, and content.