POA: Pre-training Once for Models of All Sizes
Open paper
Benchmark
Model
Rank
Results
object-detection-on-coco-val2017
POA ViT-B/16 (Cascade Mask R-CNN)
#113
box AP: 52.4
object-detection-on-coco-val2017
POA ViT-S/16 (Cascade Mask R-CNN)
#142
box AP: 50.6
Rank counts only results with a code link.