Video AI intelligent detection method and system based on file slices

The invention discloses a video AI intelligent detection method and system based on file slices. The method comprises the following steps: receiving a live program stream; converting the live program stream into a video and audio file; delaying the video and audio file for a preset time, converting the video and audio file into a delayed program stream, and broadcasting the delayed program stream; slicing and splitting the video and audio file to obtain a plurality of video and audio slices; performing compliance detection on each video and audio slice to obtain a detection result; and if the detection result is non-compliance, generating a termination instruction, and stopping broadcasting the delayed program stream according to the termination instruction. According to the invention, under the condition that the playing quality is ensured, the situation that illegal contents are missed to report is reduced. 本发明公开了一种基于文件切片的视频AI智能检测方法及系统，本方法通过接收直播节目流；将直播节目流转换为视音频文件；对视音频文件延时预设时间后，转换为延时后节目流，并播送延时后节目流；对视音频文件进行切片