Tracking the Copyright of Large Vision-Language Models through Parameter Learning Adversarial Images

Open in new window