Introducing Mask Grounding improves visual grounding in language features, enhancing Referring Image Segmentation algorithms.
