Pix2Cap-COCO: Advancing Visual Comprehension via Pixel-Level Captioning

Open in new window