The Potential of Vision-Language Models for Content Moderation of Children's Videos | IEEE Conference Publication | IEEE Xplore