SegFormer B5 ADE20k

JSON →
nvidia vision
image

SegFormer B5 is a semantic segmentation model fine-tuned on ADE20k at 640x640 resolution.

vision