NYU-Depth-v2 dataset labeled many objects. I only use 40 classes, but there are still too many bboxes.
After training, the test still produces an image with the same shape as the boxes.
prompt: In the dense summer forest, two cats and a dog play while a monkey swings through the branches.
NYU-Depth-v2 dataset labeled many objects. I only use 40 classes, but there are still too many bboxes. After training, the test still produces an image with the same shape as the boxes. prompt: In the dense summer forest, two cats and a dog play while a monkey swings through the branches.