centernet_swint.py

_base_ = '../_base_/default_runtime.py'
pretrained = 'https://github.com/SwinTransformer/storage/releases/download/v1.0.0/swin_tiny_patch4_window7_224.pth'  # noqa
model = dict(
    type='CenterNet',
    backbone=dict(
        type='SwinTransformer',
        embed_dims=96,
        depths=[2, 2, 6, 2],
        num_heads=[3, 6, 12, 24],
        window_size=7,
        mlp_ratio=4,
        qkv_bias=True,
        qk_scale=None,
        drop_rate=0.,
        attn_drop_rate=0.,
        drop_path_rate=0.2,
        patch_norm=True,
        out_indices=(0, 1, 2, 3),
        # Please only add indices that would be used
        # in FPN, otherwise some parameter will not be used
        with_cp=False,
        convert_weights=True,
        init_cfg=dict(type='Pretrained', checkpoint=pretrained)),
    neck=dict(
        type='CTNeck_SR',
        in_channel=768,
        channel_dims=[384, 192, 96]),
    bbox_head=dict(
        type='CenterNetHead',
        num_classes=80,
        in_channel=96,
        feat_channel=96,
        loss_center_heatmap=dict(type='GaussianFocalLoss', loss_weight=1.0),
        loss_wh=dict(type='L1Loss', loss_weight=0.1),
        loss_offset=dict(type='L1Loss', loss_weight=1.0)),
    train_cfg=None,
    test_cfg=dict(topk=100, local_maximum_kernel=3, max_per_img=100))

dataset_type = 'CocoDataset'
data_root = 'data/coco/'
# We fixed the incorrect img_norm_cfg problem in the source code.
img_norm_cfg = dict(
    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)

train_pipeline = [
    dict(type='LoadImageFromFile', to_float32=True, color_type='color'),
    dict(type='LoadAnnotations', with_bbox=True),
    dict(
        type='RandomCenterCropPad',
        crop_size=(512, 512),
        ratios=(0.6, 0.7, 0.8, 0.9, 1.0, 1.1, 1.2, 1.3),
        mean=[0, 0, 0],
        std=[1, 1, 1],
        to_rgb=True,
        test_pad_mode=None),
    dict(type='Resize', img_scale=(512, 512), keep_ratio=True),
    dict(type='RandomFlip', flip_ratio=0.5),
    dict(type='Normalize', **img_norm_cfg),
    dict(type='DefaultFormatBundle'),
    dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels'])
]

test_pipeline = [
    dict(type='LoadImageFromFile', to_float32=True),
    dict(
        type='MultiScaleFlipAug',
        scale_factor=1.0,
        flip=False,
        transforms=[
            dict(type='Resize', keep_ratio=True),
            dict(
                type='RandomCenterCropPad',
                ratios=None,
                border=None,
                mean=[0, 0, 0],
                std=[1, 1, 1],
                to_rgb=True,
                test_mode=True,
                test_pad_mode=['logical_or', 31],
                test_pad_add_pix=1),
            dict(type='RandomFlip'),
            dict(type='Normalize', **img_norm_cfg),
            dict(type='DefaultFormatBundle'),
            dict(
                type='Collect',
                meta_keys=('filename', 'ori_shape', 'img_shape', 'pad_shape',
                           'scale_factor', 'flip', 'flip_direction',
                           'img_norm_cfg', 'border'),
                keys=['img'])
        ])
]

# Use RepeatDataset to speed up training
data = dict(
    samples_per_gpu=16,
    workers_per_gpu=4,
    train=dict(
        type='RepeatDataset',
        times=5,
        dataset=dict(
            type=dataset_type,
            ann_file=data_root + 'annotations/instances_train2017.json',
            img_prefix=data_root + 'train2017/',
            pipeline=train_pipeline)),
    val=dict(
        type=dataset_type,
        ann_file=data_root + 'annotations/instances_val2017_down1_4.json',
        img_prefix=data_root + 'val2017_down1_4/',
        pipeline=test_pipeline),
    test=dict(
        type=dataset_type,
        ann_file=data_root + 'annotations/instances_val2017_down1_4.json',
        img_prefix=data_root + 'val2017_down1_4/',
        pipeline=test_pipeline))


# lr =0.01 for batch size 4
optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0001)
optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))

# learning policy
# Based on the default settings of modern detectors, we added warmup settings.
lr_config = dict(
    policy='step',
    warmup='linear',
    warmup_iters=1000,
    warmup_ratio=1.0 / 1000,
    step=[18, 24])  # the real step is [18*5, 24*5]
runner = dict(type='EpochBasedRunner', max_epochs=28)  # the real epoch is 28*5=140
evaluation = dict(interval=1, metric='bbox')
find_unused_parameters = True