[New Configs] Add new config file for MobileNetV3 and cityscapes in mmseg/configs folder

mmseg/configs/_base_/datasets/cityscapes.py

@@ -0,0 +1,79 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+from mmcv.transforms.loading import LoadImageFromFile
+from mmcv.transforms.processing import (RandomFlip, RandomResize, Resize,
+                                        TestTimeAug)
+from mmengine.dataset.sampler import DefaultSampler, InfiniteSampler
+
+from mmseg.datasets.cityscapes import CityscapesDataset
+from mmseg.datasets.transforms.formatting import PackSegInputs
+from mmseg.datasets.transforms.loading import LoadAnnotations
+from mmseg.datasets.transforms.transforms import (PhotoMetricDistortion,
+                                                  RandomCrop)
+from mmseg.evaluation import IoUMetric
+
+# dataset settings
+dataset_type = CityscapesDataset
+data_root = 'data/cityscapes/'
+crop_size = (512, 1024)
+train_pipeline = [
+    dict(type=LoadImageFromFile),
+    dict(type=LoadAnnotations),
+    dict(
+        type=RandomResize,
+        scale=(2048, 1024),
+        ratio_range=(0.5, 2.0),
+        keep_ratio=True),
+    dict(type=RandomCrop, crop_size=crop_size, cat_max_ratio=0.75),
+    dict(type=RandomFlip, prob=0.5),
+    dict(type=PhotoMetricDistortion),
+    dict(type=PackSegInputs)
+]
+test_pipeline = [
+    dict(type=LoadImageFromFile),
+    dict(type=Resize, scale=(2048, 1024), keep_ratio=True),
+    # add loading annotation after ``Resize`` because ground truth
+    # does not need to do resize data transform
+    dict(type=LoadAnnotations),
+    dict(type=PackSegInputs)
+]
+img_ratios = [0.5, 0.75, 1.0, 1.25, 1.5, 1.75]
+tta_pipeline = [
+    dict(type=LoadImageFromFile, backend_args=None),
+    dict(
+        type=TestTimeAug,
+        transforms=[[
+            dict(type=Resize, scale_factor=r, keep_ratio=True)
+            for r in img_ratios
+        ],
+                    [
+                        dict(type=RandomFlip, prob=0., direction='horizontal'),
+                        dict(type=RandomFlip, prob=1., direction='horizontal')
+                    ], [dict(type=LoadAnnotations)],
+                    [dict(type=PackSegInputs)]])
+]
+train_dataloader = dict(
+    batch_size=2,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type=InfiniteSampler, shuffle=True),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_prefix=dict(
+            img_path='leftImg8bit/train', seg_map_path='gtFine/train'),
+        pipeline=train_pipeline))
+val_dataloader = dict(
+    batch_size=1,
+    num_workers=4,
+    persistent_workers=True,
+    sampler=dict(type=DefaultSampler, shuffle=False),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_prefix=dict(
+            img_path='leftImg8bit/val', seg_map_path='gtFine/val'),
+        pipeline=test_pipeline))
+test_dataloader = val_dataloader
+
+val_evaluator = dict(type=IoUMetric, iou_metrics=['mIoU'])
+test_evaluator = val_evaluator

mmseg/configs/_base_/models/lraspp_m_v3_d8.py

@@ -0,0 +1,43 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+from torch.nn.modules.activation import ReLU
+from torch.nn.modules.batchnorm import SyncBatchNorm as SyncBN
+
+from mmseg.models.backbones import MobileNetV3
+from mmseg.models.data_preprocessor import SegDataPreProcessor
+from mmseg.models.decode_heads import LRASPPHead
+from mmseg.models.losses import CrossEntropyLoss
+from mmseg.models.segmentors import EncoderDecoder
+
+# model settings
+norm_cfg = dict(type=SyncBN, eps=0.001, requires_grad=True)
+data_preprocessor = dict(
+    type=SegDataPreProcessor,
+    mean=[123.675, 116.28, 103.53],
+    std=[58.395, 57.12, 57.375],
+    bgr_to_rgb=True,
+    pad_val=0,
+    seg_pad_val=255)
+model = dict(
+    type=EncoderDecoder,
+    data_preprocessor=data_preprocessor,
+    backbone=dict(
+        type=MobileNetV3,
+        arch='large',
+        out_indices=(1, 3, 16),
+        norm_cfg=norm_cfg),
+    decode_head=dict(
+        type=LRASPPHead,
+        in_channels=(16, 24, 960),
+        in_index=(0, 1, 2),
+        channels=128,
+        input_transform='multiple_select',
+        dropout_ratio=0.1,
+        num_classes=19,
+        norm_cfg=norm_cfg,
+        act_cfg=dict(type=ReLU),
+        align_corners=False,
+        loss_decode=dict(
+            type=CrossEntropyLoss, use_sigmoid=False, loss_weight=1.0)),
+    # model training and testing settings
+    train_cfg=dict(),
+    test_cfg=dict(mode='whole'))

mmseg/configs/mobilenet_v3/mobilenet_v3_d8_lraspp_4xb4_320k_cityscapes_512x1024.py

@@ -0,0 +1,22 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+from mmengine.config import read_base
+from mmengine.model.weight_init import PretrainedInit
+
+with read_base():
+    from .._base_.datasets.cityscapes import *
+    from .._base_.default_runtime import *
+    from .._base_.models.lraspp_m_v3_d8 import *
+    from .._base_.schedules.schedule_320k import *
+
+checkpoint = 'open-mmlab://contrib/mobilenet_v3_large'
+crop_size = (512, 1024)
+data_preprocessor.update(dict(size=crop_size))
+model.update(
+    dict(
+        data_preprocessor=data_preprocessor,
+        backbone=dict(
+            init_cfg=dict(type=PretrainedInit, checkpoint=checkpoint))))
+# Re-config the data sampler.
+train_dataloader.update(dict(batch_size=4, num_workers=4))
+val_dataloader.update(dict(batch_size=1, num_workers=4))
+test_dataloader = val_dataloader

mmseg/configs/mobilenet_v3/mobilenet_v3_d8_s_lraspp_4xb4_320k_cityscapes_512x1024.py

@@ -0,0 +1,30 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+from mmengine.config import read_base
+
+with read_base():
+    from .mobilenet_v3_d8_lraspp_4xb4_320k_cityscapes_512x1024 import *
+
+checkpoint = 'open-mmlab://contrib/mobilenet_v3_small'
+norm_cfg.update(dict(type=SyncBN, eps=0.001, requires_grad=True))
+model.update(
+    dict(
+        type=EncoderDecoder,
+        backbone=dict(
+            type=MobileNetV3,
+            init_cfg=dict(type=PretrainedInit, checkpoint=checkpoint),
+            arch='small',
+            out_indices=(0, 1, 12),
+            norm_cfg=norm_cfg),
+        decode_head=dict(
+            type=LRASPPHead,
+            in_channels=(16, 16, 576),
+            in_index=(0, 1, 2),
+            channels=128,
+            input_transform='multiple_select',
+            dropout_ratio=0.1,
+            num_classes=19,
+            norm_cfg=norm_cfg,
+            act_cfg=dict(type=ReLU),
+            align_corners=False,
+            loss_decode=dict(
+                type=CrossEntropyLoss, use_sigmoid=False, loss_weight=1.0))))

mmseg/configs/mobilenet_v3/mobilenet_v3_d8_scratch_lraspp_4xb4_320k_cityscapes_512x1024.py

@@ -0,0 +1,16 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+from mmengine.config import read_base
+
+with read_base():
+    from .._base_.datasets.cityscapes import *
+    from .._base_.default_runtime import *
+    from .._base_.models.lraspp_m_v3_d8 import *
+    from .._base_.schedules.schedule_320k import *
+
+crop_size = (512, 1024)
+data_preprocessor = dict(size=crop_size)
+# Re-config the data sampler.
+model.update(dict(data_preprocessor=data_preprocessor))
+train_dataloader.update(dict(batch_size=4, num_workers=4))
+val_dataloader.update(dict(batch_size=1, num_workers=4))
+test_dataloader = val_dataloader

mmseg/configs/mobilenet_v3/mobilenet_v3_d8_scratch_s_lraspp_4xb4_320k_cityscapes_512x1024.py

@@ -0,0 +1,28 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+from mmengine.config import read_base
+
+with read_base():
+    from .mobilenet_v3_d8_scratch_lraspp_4xb4_320k_cityscapes_512x1024 import *
+
+norm_cfg.update(dict(type=SyncBN, eps=0.001, requires_grad=True))
+model.update(
+    dict(
+        type=EncoderDecoder,
+        backbone=dict(
+            type=MobileNetV3,
+            arch='small',
+            out_indices=(0, 1, 12),
+            norm_cfg=norm_cfg),
+        decode_head=dict(
+            type=LRASPPHead,
+            in_channels=(16, 16, 576),
+            in_index=(0, 1, 2),
+            channels=128,
+            input_transform='multiple_select',
+            dropout_ratio=0.1,
+            num_classes=19,
+            norm_cfg=norm_cfg,
+            act_cfg=dict(type=ReLU),
+            align_corners=False,
+            loss_decode=dict(
+                type=CrossEntropyLoss, use_sigmoid=False, loss_weight=1.0))))