Soft Teacher + Swin-L(HTC++, multi-scale)

Reported on 5 benchmarks across 3 tasks · 1 paper · 5 SOTA

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

Computer Vision3 results

Instance SegmentationonCOCO minival
mask AP· uses extra data· 2021-06-16
52.5
best: 56.6 (Co-DETR)
SOTA
End-to-End Semi-Supervised Object Detection with Soft Teacher arXiv:2106.09018
Semi-Supervised Object DetectiononCOCO 5% labeled data
mAP· 2021-06-16
30.74
best: 40.1 (Semi-DETR)
SOTA
End-to-End Semi-Supervised Object Detection with Soft Teacher arXiv:2106.09018
Semi-Supervised Object DetectiononCOCO 1% labeled data
mAP· 2021-06-16
20.46
best: 31.7 (MixPL)
SOTA
End-to-End Semi-Supervised Object Detection with Soft Teacher arXiv:2106.09018

Methodology2 results

2D Object DetectiononCOCO 5% labeled data
mAP· 2021-06-16
30.74
best: 40.1 (Semi-DETR)
SOTA
End-to-End Semi-Supervised Object Detection with Soft Teacher arXiv:2106.09018
2D Object DetectiononCOCO 1% labeled data
mAP· 2021-06-16
20.46
best: 31.7 (MixPL)
SOTA
End-to-End Semi-Supervised Object Detection with Soft Teacher arXiv:2106.09018