16k on TBBR

Metric: Average Recall@IoU:0.5-0.95 (higher is better)

LeaderboardDataset

Results

Sort:

#	Model↕	Average Recall@IoU:0.5-0.95▼	Augmentations	Paper	Date↕	Code
1	Swin-T (ImageNet-1k pretrain)	45.4	No	-	-	Code
2	FSAF (ResNeXt-101, ImageNet-1k pretrain)	38	No	-	-	Code
3	Mask R-CNN (ResNet-50-FPN, ImageNet-1k pretrain)	37	No	-	-	Code
4	Mask R-CNN (ResNet-50-FPN)	30.8	No	-	-	Code
5	TridentNet (ResNet-50, ImageNet-1k pretrain)	30	No	-	-	Code
6	FSAF (ResNeXt-101)	24.8	No	-	-	Code
7	TridentNet (ResNet-50)	21.5	No	-	-	Code

#1Swin-T (ImageNet-1k pretrain)
45.4
Average Recall@IoU:0.5-0.95
No paperCode
#2FSAF (ResNeXt-101, ImageNet-1k pretrain)
38
Average Recall@IoU:0.5-0.95
No paperCode
#3Mask R-CNN (ResNet-50-FPN, ImageNet-1k pretrain)
37
Average Recall@IoU:0.5-0.95
No paperCode
#4Mask R-CNN (ResNet-50-FPN)
30.8
Average Recall@IoU:0.5-0.95
No paperCode
#5TridentNet (ResNet-50, ImageNet-1k pretrain)
30
Average Recall@IoU:0.5-0.95
No paperCode
#6FSAF (ResNeXt-101)
24.8
Average Recall@IoU:0.5-0.95
No paperCode
#7TridentNet (ResNet-50)
21.5
Average Recall@IoU:0.5-0.95
No paperCode