SynRES: Towards Referring Expression Segmentation in the Wild via Synthetic Data

Open in new window