Method for detecting mobile phone answering and dialing behaviors of user based on deep learning multi-model fusion

The invention discloses a user mobile phone answering and beating behavior detection method based on deep learning multi-model fusion. The method comprises the following steps: obtaining a person hand raising action probability and a probability that a mobile phone is located at the head of a human body; and inputting results of the person hand-raising action discriminator and the mobile phone discriminator into the person answering and hitting discriminator, and finally determining whether a person answers and hits the mobile phone in the target video stream according to the person hand-raising action probability and the probability that the mobile phone is located at the head of the human body. The prediction results of the mobile phone detection model and the human body key point detection model are integrated, and when the calculation formula prediction probability of any one model is greater than the set value, it is directly judged that a person answers and makes a mobile phone, so that the situation th