Abstract: Unlike traditional image or video steganography in spatial/transform domain, motion vector (MV)-based methods target the internal dynamics of video compression and embed messages while ...