Panoptic Segmentation on COCO minival

Metric: SQ (higher is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Sort:

#	Model↕	SQ▼	Extra Data	Paper	Date↕	Code
1	Panoptic FCN* (Swin-L, single-scale)	83.2	No	Fully Convolutional Networks for Panoptic Segmen...	2020-12-01	Code
2	MaskFormer (single-scale)	81.8	No	Per-Pixel Classification is Not All You Need for...	2021-07-13	Code
3	Panoptic FCN* (ResNet-50-FPN)	80.7	No	Fully Convolutional Networks for Panoptic Segmen...	2020-12-01	Code
4	DETR-R101 (ResNet-101)	79.9	No	End-to-End Object Detection with Transformers	2020-05-26	Code
5	PanopticFPN++	79.5	No	End-to-End Object Detection with Transformers	2020-05-26	Code

#1Panoptic FCN* (Swin-L, single-scale)SOTA
83.2
SQ· 2020-12-01
Fully Convolutional Networks for Panoptic Segmentation Code
#2MaskFormer (single-scale)
81.8
SQ· 2021-07-13
Per-Pixel Classification is Not All You Need for Semantic Segmentation Code
#3Panoptic FCN* (ResNet-50-FPN)
80.7
SQ· 2020-12-01
Fully Convolutional Networks for Panoptic Segmentation Code
#4DETR-R101 (ResNet-101)SOTA
79.9
SQ· 2020-05-26
End-to-End Object Detection with Transformers Code
#5PanopticFPN++
79.5
SQ· 2020-05-26
End-to-End Object Detection with Transformers Code