Closed dqj5182 closed 8 months ago
Which Mask2Former pretrained checkpoint did the paper use to conduct scene segmentation? (among https://github.com/facebookresearch/Mask2Former/blob/main/MODEL_ZOO.md)
Was it custom trained and tested on the DECO, RICH, PROX dataset? or was it pretrained models of segmentation dataset (i.e., Cityscape)?
A pretrained model was used. More specifically, this config.
Which Mask2Former pretrained checkpoint did the paper use to conduct scene segmentation? (among https://github.com/facebookresearch/Mask2Former/blob/main/MODEL_ZOO.md)
Was it custom trained and tested on the DECO, RICH, PROX dataset? or was it pretrained models of segmentation dataset (i.e., Cityscape)?