SegFormer B1 ADE20k

JSON →
nvidia vision
image

SegFormer B1 is a semantic segmentation model fine-tuned on ADE20k at 512x512 resolution.

vision