Fetching the paper…

VLM: Task-agnostic Video-Language Model Pre-training for Video Understanding · Around