Official PyTorch implementation of ODISE: Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models [CVPR 2023 Highlight]
845
stars
45
forks
source link
512x512 configuration as in ablation studies #16
Open
volgachen opened 1 year ago
Hello, could you share the $512\times512$ configuration used in the ablation study? Is there any other change other than the resolution?
I've just modify all
1024
into512
in configs/common/data/coco_panoptic_semseg.py. It diffs like this:I suppose $512\times512$ does not require slides so I turned it off as well. I wonder if these are consistent with your configuration.