ViT-CoMer: Vision Transformer with Convolutional Multi-scale Feature Interaction for Dense Predictions
Open paper
Other version
Benchmark
Model
Rank
Results
instance-segmentation-on-coco-minival
ViT-CoMer-L (Mask RCNN, DINOv2)
#2
mask AP: 55.9
semantic-segmentation-on-ade20k-val
ViT-CoMer
#2
mIoU: 62.1
object-detection-on-coco-val2017
ViT-CoMer
#9
box AP: 64.3
Params (M): 363
Rank counts only results with a code link.