semantic segmentation paper list

A paper list of semantic segmentation.

*Last updated: 2022/02/15

Update log

2022/January - update some recent papers (TITS, PR, IEEE SPL, arXiv) of semantic segmentation.

Table of Contents

• Datasets
• Metrics
• Performance tables
• Paper list

Datasets

The papers related to datasets used mainly in natural/color image segmentation are as follows.

• [Cityscapes] Cityscapes is a large-scale database which focuses on semantic understanding of urban street scenes. It provides semantic, instance-wise, and dense pixel annotations for 30 classes grouped into 8 categories (flat surfaces, humans, vehicles, constructions, objects, nature, sky, and void).
• [Pascal VOC] PASCAL VOC 2007 is a dataset for image recognition and image semantic segmentation.
• [ADE20K] The ADE20K semantic segmentation dataset contains more than 20K scene-centric images exhaustively annotated with pixel-level objects and object parts labels. There are totally 150 semantic categories, which include stuffs like sky, road, grass, and discrete objects like person, car, bed.

Metrics

The papers related to metrics used mainly in RGBD semantic segmentation are as follows.

• [PixAcc] Pixel accuracy
• [mAcc] Mean accuracy
• [mIoU] Mean intersection over union
• [f.w.IOU] Frequency weighted IOU

Performance tables

We show several evaluation indicators and backbones used in the performance table, and indicate the year and source of the article.

Performance tables

We show several evaluation indicators and backbones used in the performance table, and indicate the year and source of the article.

Cityscapes

Method	Backbone	PixAcc	mAcc	mIoU(val)	mIoU(test)	f.w.IOU	Published	Year
PSPNet	ResNet101	-	-	-	78.4	-	CVPR	2017
BiseNet.V1	ResNet101			80.3	78.9		ECCV	2018
BiseNet.V2-L		-	-	75.8	75.3	-		2020
STDC	STDC2			77.0	76.8		CVPR	2021

CamVid

Method	Backbone	PixAcc	mAcc	mIoU	f.w.IOU	Published	Year
PSPNet	ResNet50	-	-	69.1	-	CVPR	2017
BiseNet.V1	ResNet18			68.7	-	ECCV	2018
BiseNet.V2-L	-			73.2			2020

ADE20K

Method	Backbone	PixAcc	mAcc	mIoU	f.w.IOU	Published	Year
PSPNet	ResNet101	-	-	-	-	CVPR	2017
BiseNet.V1
BiseNet.V2-L		-	-		-		2020
STDC	STDC2			73.9		CVPR	2021
PFD(CNN)	ResNet101			47.9		arXiv	2022
PFD(Transformer)	Swin-L			56.0		arXiv	2022

PASCAL VOC 2012

Method	PixAcc	mAcc	mIoU(val)	mIoU(test)	f.w.IOU	backbone	Published	Year
PSPNet	-	-	-	82.6	-	ResNet101	CVPR	2017
BiseNet.V2-L	-	-	75.8	75.3	-			2020

Paper list

• [PSPNet] Zhao H, Shi J, Qi X, et al. Pyramid scene parsing network[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 2881-2890. [Paper] [Code]
• [BiSeNet] Yu C, Wang J, Peng C, et al. Bisenet: Bilateral segmentation network for real-time semantic segmentation[C]//Proceedings of the European conference on computer vision (ECCV). 2018: 325-341. [Paper] [Code]
• [STDC] Fan M, Lai S, Huang J, et al. Rethinking bisenet for real-time semantic segmentation[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2021: 9716-9725. [Paper] [[Code](https://github.com/ MichaelFan01/STDC-Seg)]
• [PFD] Qin Z, Liu J, Zhang X, et al. Pyramid Fusion Transformer for Semantic Segmentation[J]. arXiv preprint arXiv:2201.04019, 2022. [Paper] [Code]

Contact & Feedback

If you have any suggestions about this project, feel free to contact me.

• [e-mail: xiehousen[at]gmail.com]