update

.gitignore

@@ -0,0 +1,128 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+.hypothesis/
+.pytest_cache/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# pyenv
+.python-version
+
+# celery beat schedule file
+celerybeat-schedule
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+
+openselfsup/version.py
+version.py
+data
+.vscode
+.idea
+
+# custom
+*.pkl
+*.pkl.json
+*.log.json
+work_dirs/
+pretrains
+
+# Pytorch
+*.pth
+
+*.swp
+source.sh
+tensorboard.sh
+.DS_Store
+replace.sh
+benchmarks/detection/datasets
+benchmarks/detection/output

README.md

@@ -33,6 +33,10 @@
 
 
 ## Codes
+
+### Pretext Task Pretrain
+
+### Transfer to COCO
 Coming Soon
 
 ## Download Models

benchmarks/detection/README.md

@@ -0,0 +1,12 @@
+
+## Transferring to Detection
+
+We follow the evaluation setting in MoCo when trasferring to object detection.
+
+### Instruction
+
+1. Install [detectron2](https://github.com/facebookresearch/detectron2/blob/master/INSTALL.md).
+
+1. Put dataset under "benchmarks/detection/datasets" directory,
+   following the [directory structure](https://github.com/facebookresearch/detectron2/tree/master/datasets)
+	 requried by detectron2.

benchmarks/detection/configs/Base-Keypoint-RCNN-FPN.yaml

@@ -0,0 +1,15 @@
+_BASE_: "Base-RCNN-FPN.yaml"
+MODEL:
+  KEYPOINT_ON: True
+  ROI_HEADS:
+    NUM_CLASSES: 1
+  ROI_BOX_HEAD:
+    SMOOTH_L1_BETA: 0.5  # Keypoint AP degrades (though box AP improves) when using plain L1 loss
+  RPN:
+    # Detectron1 uses 2000 proposals per-batch, but this option is per-image in detectron2.
+    # 1000 proposals per-image is found to hurt box AP.
+    # Therefore we increase it to 1500 per-image.
+    POST_NMS_TOPK_TRAIN: 1500
+DATASETS:
+  TRAIN: ("keypoints_coco_2017_train",)
+  TEST: ("keypoints_coco_2017_val",)

benchmarks/detection/configs/Base-RCNN-C4-BN.yaml

@@ -0,0 +1,17 @@
+MODEL:
+  META_ARCHITECTURE: "GeneralizedRCNN"
+  RPN:
+    PRE_NMS_TOPK_TEST: 6000
+    POST_NMS_TOPK_TEST: 1000
+  ROI_HEADS:
+    NAME: "Res5ROIHeadsExtraNorm"
+  BACKBONE:
+    FREEZE_AT: 0
+  RESNETS:
+    NORM: "SyncBN"
+TEST:
+  PRECISE_BN:
+    ENABLED: True
+SOLVER:
+  IMS_PER_BATCH: 16
+  BASE_LR: 0.02

benchmarks/detection/configs/Base-RCNN-FPN.yaml

@@ -0,0 +1,42 @@
+MODEL:
+  META_ARCHITECTURE: "GeneralizedRCNN"
+  BACKBONE:
+    NAME: "build_resnet_fpn_backbone"
+  RESNETS:
+    OUT_FEATURES: ["res2", "res3", "res4", "res5"]
+  FPN:
+    IN_FEATURES: ["res2", "res3", "res4", "res5"]
+  ANCHOR_GENERATOR:
+    SIZES: [[32], [64], [128], [256], [512]]  # One size for each in feature map
+    ASPECT_RATIOS: [[0.5, 1.0, 2.0]]  # Three aspect ratios (same for all in feature maps)
+  RPN:
+    IN_FEATURES: ["p2", "p3", "p4", "p5", "p6"]
+    PRE_NMS_TOPK_TRAIN: 2000  # Per FPN level
+    PRE_NMS_TOPK_TEST: 1000  # Per FPN level
+    # Detectron1 uses 2000 proposals per-batch,
+    # (See "modeling/rpn/rpn_outputs.py" for details of this legacy issue)
+    # which is approximately 1000 proposals per-image since the default batch size for FPN is 2.
+    POST_NMS_TOPK_TRAIN: 1000
+    POST_NMS_TOPK_TEST: 1000
+  ROI_HEADS:
+    NAME: "StandardROIHeads"
+    IN_FEATURES: ["p2", "p3", "p4", "p5"]
+  ROI_BOX_HEAD:
+    NAME: "FastRCNNConvFCHead"
+    NUM_FC: 2
+    POOLER_RESOLUTION: 7
+  ROI_MASK_HEAD:
+    NAME: "MaskRCNNConvUpsampleHead"
+    NUM_CONV: 4
+    POOLER_RESOLUTION: 14
+DATASETS:
+  TRAIN: ("coco_2017_train",)
+  TEST: ("coco_2017_val",)
+SOLVER:
+  IMS_PER_BATCH: 16
+  BASE_LR: 0.02
+  STEPS: (60000, 80000)
+  MAX_ITER: 90000
+INPUT:
+  MIN_SIZE_TRAIN: (640, 672, 704, 736, 768, 800)
+VERSION: 2

benchmarks/detection/configs/Base-RetinaNet.yaml

@@ -0,0 +1,25 @@
+MODEL:
+  META_ARCHITECTURE: "RetinaNet"
+  BACKBONE:
+    NAME: "build_retinanet_resnet_fpn_backbone"
+  RESNETS:
+    OUT_FEATURES: ["res3", "res4", "res5"]
+  ANCHOR_GENERATOR:
+    SIZES: !!python/object/apply:eval ["[[x, x * 2**(1.0/3), x * 2**(2.0/3) ] for x in [32, 64, 128, 256, 512 ]]"]
+  FPN:
+    IN_FEATURES: ["res3", "res4", "res5"]
+  RETINANET:
+    IOU_THRESHOLDS: [0.4, 0.5]
+    IOU_LABELS: [0, -1, 1]
+    SMOOTH_L1_LOSS_BETA: 0.0
+DATASETS:
+  TRAIN: ("coco_2017_train",)
+  TEST: ("coco_2017_val",)
+SOLVER:
+  IMS_PER_BATCH: 16
+  BASE_LR: 0.01  # Note that RetinaNet uses a different default learning rate
+  STEPS: (60000, 80000)
+  MAX_ITER: 90000
+INPUT:
+  MIN_SIZE_TRAIN: (640, 672, 704, 736, 768, 800)
+VERSION: 2

benchmarks/detection/configs/Cityscapes/mask_rcnn_R_50_FPN.yaml

@@ -0,0 +1,30 @@
+_BASE_: "../Base-RCNN-FPN.yaml"
+MODEL:
+  WEIGHTS: "detectron2://ImageNetPretrained/MSRA/R-50.pkl"
+  MASK_ON: True
+  ROI_HEADS:
+    NUM_CLASSES: 8
+  BACKBONE:
+    FREEZE_AT: 0
+  RESNETS:
+    DEPTH: 50
+    NORM: "SyncBN"
+  FPN:
+    NORM: "SyncBN"
+INPUT:
+  MIN_SIZE_TRAIN: (800, 832, 864, 896, 928, 960, 992, 1024)
+  MIN_SIZE_TRAIN_SAMPLING: "choice"
+  MIN_SIZE_TEST: 1024
+  MAX_SIZE_TRAIN: 2048
+  MAX_SIZE_TEST: 2048
+DATASETS:
+  TRAIN: ("cityscapes_fine_instance_seg_train",)
+  TEST: ("cityscapes_fine_instance_seg_val",)
+SOLVER:
+  BASE_LR: 0.01
+  STEPS: (18000,)
+  MAX_ITER: 24000
+  IMS_PER_BATCH: 8
+TEST:
+  PRECISE_BN:
+    ENABLED: True

benchmarks/detection/configs/Cityscapes/mask_rcnn_R_50_FPN_moco.yaml

@@ -0,0 +1,9 @@
+_BASE_: "mask_rcnn_R_50_FPN.yaml"
+MODEL:
+  PIXEL_MEAN: [123.675, 116.280, 103.530]
+  PIXEL_STD: [58.395, 57.120, 57.375]
+  WEIGHTS: "See Instructions"
+  RESNETS:
+    STRIDE_IN_1X1: False
+INPUT:
+  FORMAT: "RGB"

benchmarks/detection/configs/coco_R_50_C4_12k.yaml

@@ -0,0 +1,6 @@
+_BASE_: "coco_R_50_C4_2x.yaml"
+SOLVER:
+  STEPS: (9000, 11000)
+  MAX_ITER: 12000
+
+

benchmarks/detection/configs/coco_R_50_C4_12k_moco.yaml

@@ -0,0 +1,6 @@
+_BASE_: "coco_R_50_C4_2x_moco.yaml"
+SOLVER:
+  STEPS: (9000, 11000)
+  MAX_ITER: 12000
+
+

benchmarks/detection/configs/coco_R_50_C4_1x.yaml

@@ -0,0 +1,4 @@
+_BASE_: "coco_R_50_C4_2x.yaml"
+SOLVER:
+  STEPS: (60000, 80000)
+  MAX_ITER: 90000

benchmarks/detection/configs/coco_R_50_C4_1x_moco.yaml

@@ -0,0 +1,4 @@
+_BASE_: "coco_R_50_C4_2x_moco.yaml"
+SOLVER:
+  STEPS: (60000, 80000)
+  MAX_ITER: 90000

benchmarks/detection/configs/coco_R_50_C4_2x.yaml

@@ -0,0 +1,13 @@
+_BASE_: "Base-RCNN-C4-BN.yaml"
+MODEL:
+  MASK_ON: True
+  WEIGHTS: "detectron2://ImageNetPretrained/MSRA/R-50.pkl"
+INPUT:
+  MIN_SIZE_TRAIN: (640, 672, 704, 736, 768, 800)
+  MIN_SIZE_TEST: 800
+DATASETS:
+  TRAIN: ("coco_2017_train",)
+  TEST: ("coco_2017_val",)
+SOLVER:
+  STEPS: (120000, 160000)
+  MAX_ITER: 180000

benchmarks/detection/configs/coco_R_50_C4_2x_moco.yaml

@@ -0,0 +1,10 @@
+_BASE_: "coco_R_50_C4_2x.yaml"
+MODEL:
+  PIXEL_MEAN: [123.675, 116.280, 103.530]
+  PIXEL_STD: [58.395, 57.120, 57.375]
+  WEIGHTS: "See Instructions"
+  RESNETS:
+    STRIDE_IN_1X1: False
+INPUT:
+  MAX_SIZE_TRAIN: 1200
+  FORMAT: "RGB"

benchmarks/detection/configs/coco_R_50_FPN_12k.yaml

@@ -0,0 +1,4 @@
+_BASE_: "coco_R_50_FPN_1x.yaml"
+SOLVER:
+  STEPS: (9000, 11000)
+  MAX_ITER: 12000

benchmarks/detection/configs/coco_R_50_FPN_12k_moco.yaml

@@ -0,0 +1,4 @@
+_BASE_: "coco_R_50_FPN_1x_moco.yaml"
+SOLVER:
+  STEPS: (9000, 11000)
+  MAX_ITER: 12000

benchmarks/detection/configs/coco_R_50_FPN_1x.yaml

@@ -0,0 +1,17 @@
+_BASE_: "Base-RCNN-FPN.yaml"
+MODEL:
+  MASK_ON: True
+  WEIGHTS: "detectron2://ImageNetPretrained/MSRA/R-50.pkl"
+  BACKBONE:
+    FREEZE_AT: 0
+  RESNETS:
+    DEPTH: 50
+    NORM: "SyncBN"
+  FPN:
+    NORM: "SyncBN"
+TEST:
+  PRECISE_BN:
+    ENABLED: True
+SOLVER:
+  STEPS: (60000, 80000)
+  MAX_ITER: 90000

benchmarks/detection/configs/coco_R_50_FPN_1x_moco.yaml

@@ -0,0 +1,9 @@
+_BASE_: "coco_R_50_FPN_1x.yaml"
+MODEL:
+  PIXEL_MEAN: [123.675, 116.280, 103.530]
+  PIXEL_STD: [58.395, 57.120, 57.375]
+  WEIGHTS: "See Instructions"
+  RESNETS:
+    STRIDE_IN_1X1: False
+INPUT:
+  FORMAT: "RGB"

benchmarks/detection/configs/coco_R_50_FPN_2x.yaml

@@ -0,0 +1,4 @@
+_BASE_: "coco_R_50_FPN_1x.yaml"
+SOLVER:
+  STEPS: (120000, 160000)
+  MAX_ITER: 180000