deeplabv3

Rethinking Atrous Convolution for Semantic Image Segmentation

Reference

Chen, Liang-Chieh, George Papandreou, Florian Schroff, and Hartwig Adam. "Rethinking Atrous Convolution for Semantic Image Segmentation." arXiv preprint arXiv:1706.05587 (2017).

Performance

Cityscapes

Model	Backbone	Resolution	Training Iters	mIoU	mIoU (flip)	mIoU (ms+flip)	Links
DeepLabV3	ResNet50_OS8	1024x512	80000	79.90%	80.22%	80.89%	model | log | vdl
DeepLabV3	ResNet101_OS8	1024x512	80000	80.85%	81.09%	81.78%	model | log | vdl

Pascal VOC 2012 + Aug

Model	Backbone	Resolution	Training Iters	mIoU	mIoU (flip)	mIoU (ms+flip)	Links
DeepLabV3	ResNet50_OS8	512x512	40000	79.76%	80.11%	81.05%	model | log | vdl
DeepLabV3	ResNet101_OS8	512x512	40000	80.62%	80.87%	81.59%	model | log | vdl