Clean train set

configs/_base_/datasets/ade20k.py

@@ -1,6 +1,6 @@
 # dataset settings
 dataset_type = 'ADE20KDataset'
-data_root = 'data/ade/ADEChallengeData2016'
+data_root = "/media/ids/Ubuntu files/data/ADEChallengeData2016/"
 crop_size = (512, 512)
 train_pipeline = [
     dict(type='LoadImageFromFile'),

configs/_base_/datasets/hots_v1_640x480.py

@@ -0,0 +1,73 @@
+dataset_type = "HOTSDataset"
+data_root = "/media/ids/Ubuntu files/data/HOTS_v1/SemanticSegmentation/"
+crop_size = (640, 480)
+train_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(type='LoadAnnotations'),
+    dict(
+        type='RandomResize',
+        scale=(2048, 480),
+        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='PackSegInputs')
+]
+test_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(type='Resize', scale=(2048, 480), 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='img_dir/train', seg_map_path='ann_dir/train'),
+        pipeline=train_pipeline))
+val_dataloader = dict(
+    batch_size=1,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_prefix=dict(img_path='img_dir/eval', seg_map_path='ann_dir/eval'),
+        pipeline=test_pipeline))
+test_dataloader = dict(
+    batch_size=1,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_prefix=dict(img_path='img_dir/test', seg_map_path='ann_dir/test'),
+        pipeline=test_pipeline))
+
+val_evaluator = dict(type='IoUMetric', iou_metrics=['mIoU'])
+test_evaluator = val_evaluator

configs/_base_/datasets/irl_vision_sim_512x512.py

@@ -0,0 +1,73 @@
+dataset_type = "IRLVisionSimDataset"
+data_root = "/media/ids/Ubuntu files/data/irl_vision_sim/SemanticSegmentation/"
+crop_size = (512, 512)
+train_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(type='LoadAnnotations'),
+    dict(
+        type='RandomResize',
+        scale=(2048, 512),
+        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='PackSegInputs')
+]
+test_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(type='Resize', scale=(2048, 512), 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='img_dir/train', seg_map_path='ann_dir/train'),
+        pipeline=train_pipeline))
+val_dataloader = dict(
+    batch_size=1,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_prefix=dict(img_path='img_dir/eval', seg_map_path='ann_dir/eval'),
+        pipeline=test_pipeline))
+test_dataloader = dict(
+    batch_size=1,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_prefix=dict(img_path='img_dir/test', seg_map_path='ann_dir/test'),
+        pipeline=test_pipeline))
+
+val_evaluator = dict(type='IoUMetric', iou_metrics=['mIoU'])
+test_evaluator = val_evaluator

configs/_base_/datasets/irl_vision_sim_640x480.py

@@ -0,0 +1,73 @@
+dataset_type = "IRLVisionSimDataset"
+data_root = "/media/ids/Ubuntu files/data/irl_vision_sim/SemanticSegmentation/"
+crop_size = (640, 480)
+train_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(type='LoadAnnotations'),
+    dict(
+        type='RandomResize',
+        scale=(2048, 480),
+        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='PackSegInputs')
+]
+test_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(type='Resize', scale=(2048, 480), 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='img_dir/train', seg_map_path='ann_dir/train'),
+        pipeline=train_pipeline))
+val_dataloader = dict(
+    batch_size=1,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_prefix=dict(img_path='img_dir/eval', seg_map_path='ann_dir/eval'),
+        pipeline=test_pipeline))
+test_dataloader = dict(
+    batch_size=1,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_prefix=dict(img_path='img_dir/test', seg_map_path='ann_dir/test'),
+        pipeline=test_pipeline))
+
+val_evaluator = dict(type='IoUMetric', iou_metrics=['mIoU'])
+test_evaluator = val_evaluator

configs/_base_/models/dpt_vit-b16.py

@@ -9,7 +9,7 @@
 model = dict(
     type='EncoderDecoder',
     data_preprocessor=data_preprocessor,
-    pretrained='pretrain/vit-b16_p16_224-80ecf9dd.pth', # noqa
+    # pretrained='pretrain/vit-b16_p16_224-80ecf9dd.pth', # noqa
     backbone=dict(
         type='VisionTransformer',
         img_size=224,

configs/_base_/models/fast_scnn.py

@@ -33,7 +33,8 @@
         norm_cfg=norm_cfg,
         align_corners=False,
         loss_decode=dict(
-            type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1)),
+            type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1)
+        ),
     auxiliary_head=[
         dict(
             type='FCNHead',

configs/_base_/models/upernet_vit-b16_ln_mln.py

@@ -10,7 +10,7 @@
 model = dict(
     type='EncoderDecoder',
     data_preprocessor=data_preprocessor,
-    pretrained='pretrain/jx_vit_base_p16_224-80ecf9dd.pth',
+    # pretrained='pretrain/jx_vit_base_p16_224-80ecf9dd.pth',
     backbone=dict(
         type='VisionTransformer',
         img_size=(512, 512),

configs/_base_/schedules/schedule_10k.py

@@ -0,0 +1,24 @@
+# optimizer
+optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005)
+optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer, clip_grad=None)
+# learning policy
+param_scheduler = [
+    dict(
+        type='PolyLR',
+        eta_min=1e-4,
+        power=0.9,
+        begin=0,
+        end=10000,
+        by_epoch=False)
+]
+# training schedule for 10k
+train_cfg = dict(type='IterBasedTrainLoop', max_iters=10000, val_interval=1000)
+val_cfg = dict(type='ValLoop')
+test_cfg = dict(type='TestLoop')
+default_hooks = dict(
+    timer=dict(type='IterTimerHook'),
+    logger=dict(type='LoggerHook', interval=50, log_metric_by_epoch=False),
+    param_scheduler=dict(type='ParamSchedulerHook'),
+    checkpoint=dict(type='CheckpointHook', by_epoch=False, interval=1000),
+    sampler_seed=dict(type='DistSamplerSeedHook'),
+    visualization=dict(type='SegVisualizationHook'))

configs/_base_/schedules/schedule_1k.py

@@ -0,0 +1,24 @@
+# optimizer
+optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005)
+optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer, clip_grad=None)
+# learning policy
+param_scheduler = [
+    dict(
+        type='PolyLR',
+        eta_min=1e-4,
+        power=0.9,
+        begin=0,
+        end=1000,
+        by_epoch=False)
+]
+# training schedule for 1k
+train_cfg = dict(type='IterBasedTrainLoop', max_iters=1000, val_interval=200)
+val_cfg = dict(type='ValLoop')
+test_cfg = dict(type='TestLoop')
+default_hooks = dict(
+    timer=dict(type='IterTimerHook'),
+    logger=dict(type='LoggerHook', interval=50, log_metric_by_epoch=False),
+    param_scheduler=dict(type='ParamSchedulerHook'),
+    checkpoint=dict(type='CheckpointHook', by_epoch=False, interval=200),
+    sampler_seed=dict(type='DistSamplerSeedHook'),
+    visualization=dict(type='SegVisualizationHook'))

configs/_base_/schedules/schedule_5k.py

@@ -0,0 +1,24 @@
+# optimizer
+optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005)
+optim_wrapper = dict(type='OptimWrapper', optimizer=optimizer, clip_grad=None)
+# learning policy
+param_scheduler = [
+    dict(
+        type='PolyLR',
+        eta_min=1e-4,
+        power=0.9,
+        begin=0,
+        end=5000,
+        by_epoch=False)
+]
+# training schedule for 5k
+train_cfg = dict(type='IterBasedTrainLoop', max_iters=5000, val_interval=1000)
+val_cfg = dict(type='ValLoop')
+test_cfg = dict(type='TestLoop')
+default_hooks = dict(
+    timer=dict(type='IterTimerHook'),
+    logger=dict(type='LoggerHook', interval=50, log_metric_by_epoch=False),
+    param_scheduler=dict(type='ParamSchedulerHook'),
+    checkpoint=dict(type='CheckpointHook', by_epoch=False, interval=1000),
+    sampler_seed=dict(type='DistSamplerSeedHook'),
+    visualization=dict(type='SegVisualizationHook'))

configs/convnext/convnext-tiny_upernet_8xb2-amp-10k_HOTS_v1-640x480.py

@@ -0,0 +1,54 @@
+_base_ = [
+    '../_base_/models/upernet_convnext.py', '../_base_/datasets/hots_v1_640x480.py',
+    '../_base_/default_runtime.py', '../_base_/schedules/schedule_10k.py'
+]
+crop_size = (640, 480)
+data_preprocessor = dict(size=crop_size)
+checkpoint_file = 'https://download.openmmlab.com/mmclassification/v0/convnext/downstream/convnext-tiny_3rdparty_32xb128-noema_in1k_20220301-795e9634.pth'  # noqa
+model = dict(
+    data_preprocessor=data_preprocessor,
+    backbone=dict(
+        type='mmpretrain.ConvNeXt',
+        arch='tiny',
+        out_indices=[0, 1, 2, 3],
+        drop_path_rate=0.4,
+        layer_scale_init_value=1.0,
+        gap_before_final_norm=False,
+        init_cfg=dict(
+            type='Pretrained', checkpoint=checkpoint_file,
+            prefix='backbone.')),
+    decode_head=dict(
+        in_channels=[96, 192, 384, 768],
+        num_classes=46,
+    ),
+    auxiliary_head=dict(in_channels=384, num_classes=46),
+    test_cfg=dict(mode='slide', crop_size=crop_size, stride=(341, 341)),
+)
+
+optim_wrapper = dict(
+    _delete_=True,
+    type='AmpOptimWrapper',
+    optimizer=dict(
+        type='AdamW', lr=0.0001, betas=(0.9, 0.999), weight_decay=0.05),
+    paramwise_cfg={
+        'decay_rate': 0.9,
+        'decay_type': 'stage_wise',
+        'num_layers': 6
+    },
+    constructor='LearningRateDecayOptimizerConstructor',
+    loss_scale='dynamic')
+
+param_scheduler = [
+    dict(
+        type='LinearLR', start_factor=1e-6, by_epoch=False, begin=0, end=1500),
+    dict(
+        type='PolyLR',
+        power=1.0,
+        begin=1500,
+        end=10000,
+        eta_min=0.0,
+        by_epoch=False,
+    )
+]
+
+