METHOD FOR DETECTING AUTHENTICITY OF NUMBER OF PEOPLE DURING LIVE BROADCAST, STORAGE MEDIUM, ELECTRONIC DEVICE AND SYSTEM

摘要

The invention relates to the field of network live broadcasts, and disclosed are a method for detecting authenticity of the number of people during a live broadcast, a storage medium, an electronic device and a system. The method comprises: S1: integrating a deep learning framework in a live broadcast client; S2: a live broadcast platform server periodically sending a picture containing/not containing a specific pattern to live broadcast clients of all current on-line users in a live broadcast room, and at the same time, the background of the live broadcast platform server making a mark, indicating whether a specific pattern is contained, for the picture sent to the live broadcast clients; S3: the live broadcast clients carrying out determination on the received picture, and determining whether the picture contains the specific pattern, and sending a result to the live broadcast platform server; and S4: the live broadcast platform server carrying out determination, according to the mark, on the result returned by the live broadcast clients, and if the result is correct, not carrying out processing, and if the result is wrong, stopping a corresponding live broadcast client from accessing the live broadcast platform server. According to the invention, validity of the live broadcast clients can be determined, thus ensuring fairness of the live broadcast.