List your company
ResearchFormerly WizWand

Object Detection on MS-COCO Standard 10% 2017 (val)

35.3AP50:95Dual Pseudo-label Polishing Framework
21.946425.413228.8832.3468Jun 8, 2021Aug 14, 2021Oct 20, 2021Dec 27, 2021Mar 4, 2022May 10, 2022Jul 17, 2022
Updated 8mo ago

Evaluation results

MethodLinks
Dual Pseudo-label Polishing Framework2022.0735.3
Soft TeacherRemark=ICCV 20212022.0734.04
MUMBackbone=ResNet-50, Training Protocol=COCO-Standard, Labeled Data Ratio=10%2021.1131.87
Humble TeacherRemark=CVPR 20212022.0731.61
Unbiased TeacherBackbone=ResNet-50, Training Protocol=COCO-Standard, Labeled Data Ratio=10%2021.1131.5
Unbiased TeacherRemark=ICLR 20212022.0731.5
ISMTBackbone=ResNet-50, Training Protocol=COCO-Standard, Labeled Data Ratio=10%2021.1130.53
Instant TeachingBackbone=ResNet-50, Training Protocol=COCO-Standard, Labeled Data Ratio=10%2021.1130.4
Instant-Teaching*Remark=CVPR 20212022.0730.4
Instant-TeachingRemark=CVPR 20212022.0729.45
DETRegDetector=DDETR2021.0629.12
STACBackbone=ResNet-50, Training Protocol=COCO-Standard, Labeled Data Ratio=10%2021.1128.64
STACRemark=arxiv 20202022.0728.64
SwAVDetector=DDETR2021.0627.79
Supervised2022.0726.94
ReSimDetector=DDETR2021.0626.56
SupervisedDetector=DDETR2021.0626.34
SupervisedBackbone=ResNet-50, Training Protocol=COCO-Standard, Labeled Data Ratio=10%2021.1123.86
CSDBackbone=ResNet-50, Training Protocol=COCO-Standard, Labeled Data Ratio=10%2021.1122.46