DerrickWang005 bf1bdba623 support coco stuff-10k/164k (#625)
* support coco stuff-10k/164k

* update docs

* fix docs

* update docs

* fix import lints

* Update docs/dataset_prepare.md

* Update docs/dataset_prepare.md

* Update tools/convert_datasets/coco_stuff164k.py

* Update tools/convert_datasets/coco_stuff10k.py

* Update tools/convert_datasets/coco_stuff10k.py

* Update tools/convert_datasets/coco_stuff10k.py

* Update tools/convert_datasets/coco_stuff10k.py

* Update coco_stuff.py

fix the description of the dataset

* Update dataset_prepare.md

fix the doc tree of coco stuff 10k

* Update coco_stuff10k.py

fix img_dir

* Update coco_stuff.py

fix descriptions

* Update coco_stuff164k.py

fix out_dir

* Update coco_stuff10k.py

fix save file name

* Update coco_stuff.py

fix seg_map_suffix

* Update dataset_prepare.md

fix -p

* Update dataset_prepare.md

fix doc tree

* modify coco stuff convertor

* Remove redundant code

* fix 164k convert bug

* remove redundant comment

* add deeplabv3 configs and more iterations

* replace shutil.move with shtil.copyfile

* Update deeplabv3_r50-d8_512x512_4x4_80k_coco_stuff10k.py

fix wrong config

* Update deeplabv3_r101-d8_512x512_4x4_80k_coco_stuff164k.py

fix wrong config

* fix wrong configs

* fix wrong configs

* fix wrong path for coco stuff 10k

* fix convert bugs

* fix seg_filename bug

* when nproc=0, use track progress

* rename configs: coco_stuff --> coco-stuff

* add coco-stuff 10k and 164k to README.md

* update configs

* add deeplabv3 benchmark

* add pspnet benchmark

* remove redundant comma

Co-authored-by: Junjun2016 <hejunjun@sjtu.edu.cn>
2021-09-22 20:48:08 +08:00
..
2021-09-22 20:48:08 +08:00

Rethinking atrous convolution for semantic image segmentation

Introduction

@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 and models

:::{note} D-8 here corresponding to the output stride 8 setting for DeepLab series. :::

Cityscapes

Method Backbone Crop Size Lr schd Mem (GB) Inf time (fps) mIoU mIoU(ms+flip) config download
DeepLabV3 R-50-D8 512x1024 40000 6.1 2.57 79.09 80.45 config model | log
DeepLabV3 R-101-D8 512x1024 40000 9.6 1.92 77.12 79.61 config model | log
DeepLabV3 R-50-D8 769x769 40000 6.9 1.11 78.58 79.89 config model | log
DeepLabV3 R-101-D8 769x769 40000 10.9 0.83 79.27 80.11 config model | log
DeepLabV3 R-18-D8 512x1024 80000 1.7 13.78 76.70 78.27 config model | log
DeepLabV3 R-50-D8 512x1024 80000 - - 79.32 80.57 config model | log
DeepLabV3 R-101-D8 512x1024 80000 - - 80.20 81.21 config model | log
DeepLabV3 R-18-D8 769x769 80000 1.9 5.55 76.60 78.26 config model | log
DeepLabV3 R-50-D8 769x769 80000 - - 79.89 81.06 config model | log
DeepLabV3 R-101-D8 769x769 80000 - - 79.67 80.81 config model | log
DeepLabV3 R-101-D16-MG124 512x1024 40000 4.7 - 6.96 76.71 78.63 config model | log
DeepLabV3 R-101-D16-MG124 512x1024 80000 - - 78.36 79.84 config model | log
DeepLabV3 R-18b-D8 512x1024 80000 1.6 13.93 76.26 77.88 config model | log
DeepLabV3 R-50b-D8 512x1024 80000 6.0 2.74 79.63 80.98 config model | log
DeepLabV3 R-101b-D8 512x1024 80000 9.5 1.81 80.01 81.21 config model | log
DeepLabV3 R-18b-D8 769x769 80000 1.8 5.79 76.63 77.51 config model | log
DeepLabV3 R-50b-D8 769x769 80000 6.8 1.16 78.80 80.27 config model | log
DeepLabV3 R-101b-D8 769x769 80000 10.7 0.82 79.41 80.73 config model | log

ADE20K

Method Backbone Crop Size Lr schd Mem (GB) Inf time (fps) mIoU mIoU(ms+flip) config download
DeepLabV3 R-50-D8 512x512 80000 8.9 14.76 42.42 43.28 config model | log
DeepLabV3 R-101-D8 512x512 80000 12.4 10.14 44.08 45.19 config model | log
DeepLabV3 R-50-D8 512x512 160000 - - 42.66 44.09 config model | log
DeepLabV3 R-101-D8 512x512 160000 - - 45.00 46.66 config model | log

Pascal VOC 2012 + Aug

Method Backbone Crop Size Lr schd Mem (GB) Inf time (fps) mIoU mIoU(ms+flip) config download
DeepLabV3 R-50-D8 512x512 20000 6.1 13.88 76.17 77.42 config model | log
DeepLabV3 R-101-D8 512x512 20000 9.6 9.81 78.70 79.95 config model | log
DeepLabV3 R-50-D8 512x512 40000 - - 77.68 78.78 config model | log
DeepLabV3 R-101-D8 512x512 40000 - - 77.92 79.18 config model | log

Pascal Context

Method Backbone Crop Size Lr schd Mem (GB) Inf time (fps) mIoU mIoU(ms+flip) config download
DeepLabV3 R-101-D8 480x480 40000 9.2 7.09 46.55 47.81 config model | log
DeepLabV3 R-101-D8 480x480 80000 - - 46.42 47.53 config model | log

Pascal Context 59

Method Backbone Crop Size Lr schd Mem (GB) Inf time (fps) mIoU mIoU(ms+flip) config download
DeepLabV3 R-101-D8 480x480 40000 - - 52.61 54.28 config model | log
DeepLabV3 R-101-D8 480x480 80000 - - 52.46 54.09 config model | log

COCO-Stuff 10k

Method Backbone Crop Size Lr schd Mem (GB) Inf time (fps) mIoU mIoU(ms+flip) config download
DeepLabV3 R-50-D8 512x512 20000 9.6 10.8 34.66 36.08 config model | log
DeepLabV3 R-101-D8 512x512 20000 13.2 8.7 37.30 38.42 config model | log
DeepLabV3 R-50-D8 512x512 40000 - - 35.73 37.09 config model | log
DeepLabV3 R-101-D8 512x512 40000 - - 37.81 38.80 config model | log

COCO-Stuff 164k

Method Backbone Crop Size Lr schd Mem (GB) Inf time (fps) mIoU mIoU(ms+flip) config download
DeepLabV3 R-50-D8 512x512 80000 9.6 10.8 39.38 40.03 config model | log
DeepLabV3 R-101-D8 512x512 80000 13.2 8.7 40.87 41.50 config model | log
DeepLabV3 R-50-D8 512x512 160000 - - 41.09 41.69 config model | log
DeepLabV3 R-101-D8 512x512 160000 - - 41.82 42.49 config model | log
DeepLabV3 R-50-D8 512x512 320000 - - 41.37 42.22 config model | log
DeepLabV3 R-101-D8 512x512 320000 - - 42.61 43.42 config model | log