1. Leveraging Unlabeled Data for Video Understanding using Pretrained Vision-Language Models Open Access Author: Chan, Michael Title: Leveraging Unlabeled Data for Video Understanding using Pretrained Vision-Language Models Area of Honors: Computer Science Keywords: Video UnderstandingPre-trained Vision-Language ModelsVideoBERTUnlabeled DataBERTLSTMVideo Representations File: Download Chan_Thesis_Final2.pdf Thesis Supervisors: Rui Zhang, Thesis SupervisorJohn Joseph Hannan, Thesis Honors Advisor
