Motion key point video analysis model training method and device and electronic equipment

[0078] as Figure 8 It is the flow chart of a motion key point video analysis model training method provided by the first embodiment of the application. The method comprises the following steps.

[0079] Step s801: provide the annotated video to the target key point analysis model to obtain the first result of motion key point analysis.

[0080]In this embodiment, it is used to obtain a target key point analysis model. As the premise of obtaining the target key point analysis model, first obtain the original training sample data for training the target key point analysis model. Annotated video refers to the video frame containing the target object and the key points of the human posture presented by the target object during motion have been identified and annotated. The target object can refer to the user who performs the preset movement. The preset movement can refer to sports. The human posture presented in the process of movement includes standing posture or horizontal posture. The

[0119]Corresponding to the application scenario embodiment of the motion key point video analysis model training method provided by the application and the motion key point video analysis model training method provided by the first embodiment, the second embodiment of the application also provides a motion key point video analysis model training device. Since the device embodiment is basically similar to the application scenario embodiment and the first embodiment, the description is relatively simple. For relevant parts, please refer to the partial description of the application scenario embodiment and the first embodiment. The device embodiments described below are only schematic.

[0120] Please refer to Figure 9 , which is a schematic diagram of a motion key point video analysis model training device provided in the second embodiment of the application.

[0121] The motion key point video analysis model training device includes:

[0122] A first result obtaining unit 901 for prov

[0134] In the above first embodiment, a training method of motion key point video analysis model is provided. The training method of motion key point video analysis model is mainly to train the first result of motion key point analysis obtained by the target key point analysis model based on the second result of motion key point analysis obtained by the trained teacher key point analysis model, Make the accuracy of the analysis results of the target key point analysis model on the mobile terminal equipment reach the predetermined threshold. Correspondingly, the third embodiment of the application provides another motion key point video analysis model training method. Since the motion key point video analysis model training method has been described in detail in the first embodiment, please refer to the description of the first embodiment for relevant parts, which will not be repeated here.

[0135] The third embodiment of the application provides another training method of motion key