A Tri-Layer Plugin to Improve Occluded Detection

Open paper
Benchmark Model Rank Results
instance-segmentation-on-cocoSwin-B + Cascade Mask R-CNN (tri-layer modelling)#40mask AP: 45.9
instance-segmentation-on-occluded-cocoSwin-B + Cascade Mask R-CNN (tri-layer modelling)#1Mean Recall: 63.64
instance-segmentation-on-occluded-cocoSwin-S + Mask R-CNN (tri-layer plugin)#3Mean Recall: 62.58
instance-segmentation-on-occluded-cocoSwin-T + Mask R-CNN (tri-layer plugin)#4Mean Recall: 62.00
instance-segmentation-on-separated-cocoSwin-B + Cascade Mask R-CNN (tri-layer modelling)#1Mean Recall: 36.88
instance-segmentation-on-separated-cocoSwin-S + Mask R-CNN (tri-layer plugin)#3Mean Recall: 35.80
instance-segmentation-on-separated-cocoSwin-T + Mask R-CNN (tri-layer plugin)#4Mean Recall: 34.72

Rank counts only results with a code link.