geshang/Pix2Cap-COCO download history
geshang/Pix2Cap-COCO is an image segmentation dataset on the Hugging Face Hub. In the last 30 days it was downloaded 408 times (51 in the last 7 days), and 3,727 times in total. It ranks #38,803 among datasets by monthly downloads.
Pix2Cap-COCO Dataset Description Pix2Cap-COCO is the first pixel-level captioning dataset derived from the panoptic COCO 2017 dataset, designed to provide more precise visual descriptions than traditional region-level captioning datasets. It consists of 20,550 images,