Meta-Personalizing Vision-Language Models to Find Named Instances in Video | IEEE Conference Publication | IEEE Xplore