This paper proposes V2A-Mark, a method to embed invisible visual and audio watermarks into videos to enable precise tamper localization and copyright protection. It addresses limitations of current forensic methods like poor generalizability and singular functionality. V2A-Mark combines fragile video steganography with robust watermarking, using temporal alignment and c...