4 ms·End-to-End Referring Video Object Segmentation with Multimodal Transformers1 points by EvgeniyZh 5y ago