shunk031/COCOA
COCOA dataset targets amodal segmentation, which aims to recognize and segment objects beyond their visible parts. This dataset includes labels not only for the visible parts of objects, but also for their occluded parts hidden by other objects. This enables learning to understand the full shape and position of objects.
054
download annotation if auth token exists (#11)
update CI (#9)
add notebook for demo (#8)
add flag for decoding RLE segmentation map (#3)
update README (#2)
Initialize (#1)
initial commit