BlendMask (ResNet-101 + DCN interval=3)

Reported on 6 benchmarks across 1 task · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision6 results

Instance SegmentationonCOCO test-dev
AP50· 2020-01-02
63.1
best: 80.8 (InternImage-H)
BlendMask: Top-Down Meets Bottom-Up for Instance Segmentation arXiv:2001.00309
Instance SegmentationonCOCO test-dev
AP75· 2020-01-02
44.6
best: 63.4 (Co-DETR)
BlendMask: Top-Down Meets Bottom-Up for Instance Segmentation arXiv:2001.00309
Instance SegmentationonCOCO test-dev
APL· 2020-01-02
54.5
best: 72.4 (EVA)
BlendMask: Top-Down Meets Bottom-Up for Instance Segmentation arXiv:2001.00309
Instance SegmentationonCOCO test-dev
APM· 2020-01-02
44.1
best: 60.1 (Co-DETR)
BlendMask: Top-Down Meets Bottom-Up for Instance Segmentation arXiv:2001.00309
Instance SegmentationonCOCO test-dev
APS· 2020-01-02
22.7
best: 41.6 (Co-DETR)
BlendMask: Top-Down Meets Bottom-Up for Instance Segmentation arXiv:2001.00309
Instance SegmentationonCOCO test-dev
mask AP· 2020-01-02
41.3
best: 57.1 (Co-DETR)
BlendMask: Top-Down Meets Bottom-Up for Instance Segmentation arXiv:2001.00309