Skip to content

Latest commit

 

History

History

deeplabv3

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
README.md
README.md
 
 

README.md

Introduction

Official Repo

Code Snippet

DeepLabV3 (ArXiv'2017) 
@article{chen2017rethinking,
    title={Rethinking atrous convolution for semantic image segmentation},
    author={Chen, Liang-Chieh and Papandreou, George and Schroff, Florian and Adam, Hartwig},
    journal={arXiv preprint arXiv:1706.05587},
    year={2017}
}

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 77.72% cfg | model | log
R-50-D16 ImageNet-1k-224x224 512x512 LR/POLICY/BS/EPOCH: 0.01/poly/16/60 trainaug/val 76.86% cfg | model | log
R-101-D8 ImageNet-1k-224x224 512x512 LR/POLICY/BS/EPOCH: 0.01/poly/16/60 trainaug/val 79.52% cfg | model | log
R-101-D16 ImageNet-1k-224x224 512x512 LR/POLICY/BS/EPOCH: 0.01/poly/16/60 trainaug/val 78.55% 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 43.19% cfg | model | log
R-50-D16 ImageNet-1k-224x224 512x512 LR/POLICY/BS/EPOCH: 0.01/poly/16/130 train/val 41.41% cfg | model | log
R-101-D8 ImageNet-1k-224x224 512x512 LR/POLICY/BS/EPOCH: 0.01/poly/16/130 train/val 45.16% cfg | model | log
R-101-D16 ImageNet-1k-224x224 512x512 LR/POLICY/BS/EPOCH: 0.01/poly/16/130 train/val 43.45% 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 79.62% cfg | model | log
R-50-D16 ImageNet-1k-224x224 512x1024 LR/POLICY/BS/EPOCH: 0.01/poly/8/220 train/val 78.19% cfg | model | log
R-101-D8 ImageNet-1k-224x224 512x1024 LR/POLICY/BS/EPOCH: 0.01/poly/8/220 train/val 80.28% cfg | model | log
R-101-D16 ImageNet-1k-224x224 512x1024 LR/POLICY/BS/EPOCH: 0.01/poly/8/220 train/val 78.03% cfg | model | log

PASCAL Context

Backbone Pretrain Crop Size Schedule Train/Eval Set mIoU Download
R-50-D8 ImageNet-1k-224x224 480x480 LR/POLICY/BS/EPOCH: 0.004/poly/16/260 train/val 46.31% cfg | model | log
R-101-D8 ImageNet-1k-224x224 480x480 LR/POLICY/BS/EPOCH: 0.004/poly/16/260 train/val 48.43% cfg | model | log

PASCAL Context 59

Backbone Pretrain Crop Size Schedule Train/Eval Set mIoU Download
R-50-D8 ImageNet-1k-224x224 480x480 LR/POLICY/BS/EPOCH: 0.004/poly/16/260 train/val 51.69% cfg | model | log
R-101-D8 ImageNet-1k-224x224 480x480 LR/POLICY/BS/EPOCH: 0.004/poly/16/260 train/val 53.81% cfg | model | log

More

You can also download the model weights from following sources: