Lucid.CLIP.CompLabelingRank¶
This task type uses CLIP to do component labeling, and only keep the objects with a volume (in pixels) in the specified range of ranks, where rank=1 is the largest component.
This task type uses CLIP to do component labeling, and only keep the objects with a volume (in pixels) in the specified range of ranks, where rank=1 is the largest component.
