SipMask++ (ResNet-101, single-scale test)

Reported on 6 benchmarks across 1 task · 1 paper · 2 SOTA

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

Computer Vision6 results

Instance SegmentationonMSCOCO
APL· 2020-07-29
56.8
best: 65.5 (RTMDet-Ins-x)
SOTA
SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation arXiv:2007.14772
Instance SegmentationonMSCOCO
APM· 2020-07-29
38.3
best: 48 (RTMDet-Ins-l)
SOTA
SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation arXiv:2007.14772
Instance SegmentationonMSCOCO
AP50· 2020-07-29
55.6
best: 67.4 (RTMDet-Ins-x)
SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation arXiv:2007.14772
Instance SegmentationonMSCOCO
AP75· 2020-07-29
37.6
best: 47.8 (RTMDet-Ins-x)
SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation arXiv:2007.14772
Instance SegmentationonMSCOCO
APS· 2020-07-29
11.2
best: 22.2 (RTMDet-Ins-x)
SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation arXiv:2007.14772
Instance SegmentationonMSCOCO
mask AP· 2020-07-29
35.4
best: 44.6 (RTMDet-Ins-x)
SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation arXiv:2007.14772