emanet

Introduction

Official Repo

Code Snippet

EMANet (ICCV'2019)

@inproceedings{li2019expectation,
    title={Expectation-maximization attention networks for semantic segmentation},
    author={Li, Xia and Zhong, Zhisheng and Wu, Jianlong and Yang, Yibo and Lin, Zhouchen and Liu, Hong},
    booktitle={Proceedings of the IEEE International Conference on Computer Vision},
    pages={9167--9176},
    year={2019}
}

Results

PASCAL VOC

Backbone	Pretrain	Crop Size	Schedule	Train/Eval Set	mIoU	Download
R-50-D8	ImageNet-1k-224x224	512x512	LR/POLICY/BS/EPOCH: 0.01/poly/16/60	trainaug/val	75.29%	cfg | model | log
R-50-D16	ImageNet-1k-224x224	512x512	LR/POLICY/BS/EPOCH: 0.01/poly/16/60	trainaug/val	74.86%	cfg | model | log
R-101-D8	ImageNet-1k-224x224	512x512	LR/POLICY/BS/EPOCH: 0.01/poly/16/60	trainaug/val	76.43%	cfg | model | log
R-101-D16	ImageNet-1k-224x224	512x512	LR/POLICY/BS/EPOCH: 0.01/poly/16/60	trainaug/val	76.19%	cfg | model | log

ADE20k

Backbone	Pretrain	Crop Size	Schedule	Train/Eval Set	mIoU	Download
R-50-D8	ImageNet-1k-224x224	512x512	LR/POLICY/BS/EPOCH: 0.01/poly/16/130	train/val	41.77%	cfg | model | log
R-50-D16	ImageNet-1k-224x224	512x512	LR/POLICY/BS/EPOCH: 0.01/poly/16/130	train/val	41.46%	cfg | model | log
R-101-D8	ImageNet-1k-224x224	512x512	LR/POLICY/BS/EPOCH: 0.01/poly/16/130	train/val	44.39%	cfg | model | log
R-101-D16	ImageNet-1k-224x224	512x512	LR/POLICY/BS/EPOCH: 0.01/poly/16/130	train/val	43.97%	cfg | model | log

CityScapes

Backbone	Pretrain	Crop Size	Schedule	Train/Eval Set	mIoU	Download
R-50-D8	ImageNet-1k-224x224	512x1024	LR/POLICY/BS/EPOCH: 0.01/poly/8/220	train/val	77.96%	cfg | model | log
R-50-D16	ImageNet-1k-224x224	512x1024	LR/POLICY/BS/EPOCH: 0.01/poly/8/220	train/val	76.73%	cfg | model | log
R-101-D8	ImageNet-1k-224x224	512x1024	LR/POLICY/BS/EPOCH: 0.01/poly/8/220	train/val	79.54%	cfg | model | log
R-101-D16	ImageNet-1k-224x224	512x1024	LR/POLICY/BS/EPOCH: 0.01/poly/8/220	train/val	77.54%	cfg | model | log

More

You can also download the model weights from following sources:

• BaiduNetdisk: https://pan.baidu.com/s/1gD-NJJWOtaHCtB0qHE79rA with access code s757