Video is clipped. Can you describe more how you process this?
How to define a segment of video, the content ie. in advertises, reels,... act really fast in short amount of time, which opposite with CCTV for most of the time the frames shared the same context. Beside, subjects in a scence, within a same context, they changes action which changes the meaning of the frames, how can you classify these.
Our goal is not just to clip videos through Shot Boundary Detection, but to clip videos by synthesizing various information such as object information and special effects through STT and YOLO. Therefore, we will be able to complement short time content such as commercials and reels in various ways.
