Corpus Video Moment Retrieval

The task extends the Single Video Moment Retrieval task to the Corpus setup where a single textual query is used to temporally localize relevant moments across all videos in the dataset.