CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor

Open in new window