CrossView Prompt training dataset: 315 camera pairs + captions (curated from SynCamVideo, Apache-2.0) 47ac9b1 verified Cseti commited on 21 days ago