Merge branch 'update-cfg' into 'master'

Tune performance with new schedule

See merge request open-mmlab/mmdet.3d!63

configs/benchmark/hv_pointpillars_secfpn_6x8_160e_pcdet_kitti-3d-3class.py

+# model settings
+point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
+voxel_size = [0.16, 0.16, 4]
+model = dict(
+    type='VoxelNet',
+    voxel_layer=dict(
+        max_num_points=32,  # max_points_per_voxel
+        point_cloud_range=point_cloud_range,
+        voxel_size=voxel_size,
+        max_voxels=(16000, 40000)  # (training, testing) max_coxels
+    ),
+    voxel_encoder=dict(
+        type='PillarFeatureNet',
+        in_channels=4,
+        feat_channels=[64],
+        with_distance=False,
+        voxel_size=voxel_size,
+        point_cloud_range=point_cloud_range,
+    ),
+    middle_encoder=dict(
+        type='PointPillarsScatter',
+        in_channels=64,
+        output_shape=[496, 432],
+    ),
+    backbone=dict(
+        type='SECOND',
+        in_channels=64,
+        layer_nums=[3, 5, 5],
+        layer_strides=[2, 2, 2],
+        out_channels=[64, 128, 256],
+    ),
+    neck=dict(
+        type='SECONDFPN',
+        in_channels=[64, 128, 256],
+        upsample_strides=[1, 2, 4],
+        out_channels=[128, 128, 128],
+    ),
+    bbox_head=dict(
+        type='Anchor3DHead',
+        num_classes=3,
+        in_channels=384,
+        feat_channels=384,
+        use_direction_classifier=True,
+        anchor_generator=dict(
+            type='Anchor3DRangeGenerator',
+            ranges=[
+                [0, -40.0, -0.6, 70.4, 40.0, -0.6],
+                [0, -40.0, -0.6, 70.4, 40.0, -0.6],
+                [0, -40.0, -1.78, 70.4, 40.0, -1.78],
+            ],
+            sizes=[[0.6, 0.8, 1.73], [0.6, 1.76, 1.73], [1.6, 3.9, 1.56]],
+            rotations=[0, 1.57],
+            reshape_out=False),
+        diff_rad_by_sin=True,
+        bbox_coder=dict(type='DeltaXYZWLHRBBoxCoder'),
+        loss_cls=dict(
+            type='FocalLoss',
+            use_sigmoid=True,
+            gamma=2.0,
+            alpha=0.25,
+            loss_weight=1.0),
+        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
+        loss_dir=dict(
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2),
+    ),
+)
+# model training and testing settings
+train_cfg = dict(
+    assigner=[
+        dict(  # for Pedestrian
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.5,
+            neg_iou_thr=0.35,
+            min_pos_iou=0.35,
+            ignore_iof_thr=-1),
+        dict(  # for Cyclist
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.5,
+            neg_iou_thr=0.35,
+            min_pos_iou=0.35,
+            ignore_iof_thr=-1),
+        dict(  # for Car
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.6,
+            neg_iou_thr=0.45,
+            min_pos_iou=0.45,
+            ignore_iof_thr=-1),
+    ],
+    allowed_border=0,
+    pos_weight=-1,
+    debug=False)
+test_cfg = dict(
+    use_rotate_nms=True,
+    nms_across_levels=False,
+    nms_thr=0.01,
+    score_thr=0.1,
+    min_bbox_size=0,
+    nms_pre=100,
+    max_num=50)
+
+# dataset settings
+dataset_type = 'KittiDataset'
+data_root = 'data/kitti/'
+class_names = ['Pedestrian', 'Cyclist', 'Car']
+input_modality = dict(use_lidar=True, use_camera=False)
+db_sampler = dict(
+    data_root=data_root,
+    info_path=data_root + 'kitti_dbinfos_train.pkl',
+    rate=1.0,
+    object_rot_range=[0.0, 0.0],
+    prepare=dict(
+        filter_by_difficulty=[-1],
+        filter_by_min_points=dict(
+            Car=5,
+            Pedestrian=5,
+            Cyclist=5,
+        )),
+    classes=class_names,
+    sample_groups=dict(
+        Car=15,
+        Pedestrian=10,
+        Cyclist=10,
+    ))
+
+train_pipeline = [
+    dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
+    dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
+    dict(type='ObjectSample', db_sampler=db_sampler),
+    dict(
+        type='ObjectNoise',
+        num_try=100,
+        loc_noise_std=[1.0, 1.0, 0.1],
+        global_rot_range=[0.0, 0.0],
+        rot_uniform_noise=[-0.78539816, 0.78539816]),
+    dict(type='RandomFlip3D', flip_ratio=0.5),
+    dict(
+        type='GlobalRotScale',
+        rot_uniform_noise=[-0.78539816, 0.78539816],
+        scaling_uniform_noise=[0.95, 1.05]),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='PointShuffle'),
+    dict(type='DefaultFormatBundle3D', class_names=class_names),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d']),
+]
+test_pipeline = [
+    dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(
+        type='DefaultFormatBundle3D',
+        class_names=class_names,
+        with_label=False),
+    dict(type='Collect3D', keys=['points']),
+]
+
+data = dict(
+    samples_per_gpu=6,
+    workers_per_gpu=4,
+    train=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_train.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=train_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=False),
+    val=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True),
+    test=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True))
+# optimizer
+lr = 0.001  # max learning rate
+optimizer = dict(
+    type='AdamW',
+    lr=lr,
+    betas=(0.95, 0.99),  # the momentum is change during training
+    weight_decay=0.01)
+optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))
+# learning policy
+lr_config = dict(
+    policy='cyclic',
+    target_ratio=(10, 1e-4),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+momentum_config = dict(
+    policy='cyclic',
+    target_ratio=(0.85 / 0.95, 1),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+checkpoint_config = dict(interval=1)
+evaluation = dict(interval=2)
+# yapf:disable
+log_config = dict(
+    interval=50,
+    hooks=[
+        dict(type='TextLoggerHook'),
+        dict(type='TensorboardLoggerHook')
+    ])
+# yapf:enable
+# runtime settings
+total_epochs = 160
+dist_params = dict(backend='nccl')
+log_level = 'INFO'
+work_dir = './work_dirs/pp_secfpn_80e'
+load_from = None
+resume_from = None
+workflow = [('train', 1)]

configs/benchmark/hv_pointpillars_secfpn_6x8_80e_pcdet_kitti-3d-3class.py

+# model settings
+point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
+voxel_size = [0.16, 0.16, 4]
+model = dict(
+    type='VoxelNet',
+    voxel_layer=dict(
+        max_num_points=32,  # max_points_per_voxel
+        point_cloud_range=point_cloud_range,
+        voxel_size=voxel_size,
+        max_voxels=(16000, 40000)  # (training, testing) max_coxels
+    ),
+    voxel_encoder=dict(
+        type='PillarFeatureNet',
+        in_channels=4,
+        feat_channels=[64],
+        with_distance=False,
+        voxel_size=voxel_size,
+        point_cloud_range=point_cloud_range,
+    ),
+    middle_encoder=dict(
+        type='PointPillarsScatter',
+        in_channels=64,
+        output_shape=[496, 432],
+    ),
+    backbone=dict(
+        type='SECOND',
+        in_channels=64,
+        layer_nums=[3, 5, 5],
+        layer_strides=[2, 2, 2],
+        out_channels=[64, 128, 256],
+    ),
+    neck=dict(
+        type='SECONDFPN',
+        in_channels=[64, 128, 256],
+        upsample_strides=[1, 2, 4],
+        out_channels=[128, 128, 128],
+    ),
+    bbox_head=dict(
+        type='Anchor3DHead',
+        num_classes=3,
+        in_channels=384,
+        feat_channels=384,
+        use_direction_classifier=True,
+        anchor_generator=dict(
+            type='Anchor3DRangeGenerator',
+            ranges=[
+                [0, -40.0, -0.6, 70.4, 40.0, -0.6],
+                [0, -40.0, -0.6, 70.4, 40.0, -0.6],
+                [0, -40.0, -1.78, 70.4, 40.0, -1.78],
+            ],
+            sizes=[[0.6, 0.8, 1.73], [0.6, 1.76, 1.73], [1.6, 3.9, 1.56]],
+            rotations=[0, 1.57],
+            reshape_out=False),
+        diff_rad_by_sin=True,
+        bbox_coder=dict(type='DeltaXYZWLHRBBoxCoder'),
+        loss_cls=dict(
+            type='FocalLoss',
+            use_sigmoid=True,
+            gamma=2.0,
+            alpha=0.25,
+            loss_weight=1.0),
+        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
+        loss_dir=dict(
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2),
+    ),
+)
+# model training and testing settings
+train_cfg = dict(
+    assigner=[
+        dict(  # for Pedestrian
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.5,
+            neg_iou_thr=0.35,
+            min_pos_iou=0.35,
+            ignore_iof_thr=-1),
+        dict(  # for Cyclist
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.5,
+            neg_iou_thr=0.35,
+            min_pos_iou=0.35,
+            ignore_iof_thr=-1),
+        dict(  # for Car
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.6,
+            neg_iou_thr=0.45,
+            min_pos_iou=0.45,
+            ignore_iof_thr=-1),
+    ],
+    allowed_border=0,
+    pos_weight=-1,
+    debug=False)
+test_cfg = dict(
+    use_rotate_nms=True,
+    nms_across_levels=False,
+    nms_thr=0.01,
+    score_thr=0.1,
+    min_bbox_size=0,
+    nms_pre=100,
+    max_num=50)
+
+# dataset settings
+dataset_type = 'KittiDataset'
+data_root = 'data/kitti/'
+class_names = ['Pedestrian', 'Cyclist', 'Car']
+input_modality = dict(use_lidar=True, use_camera=False)
+db_sampler = dict(
+    data_root=data_root,
+    info_path=data_root + 'kitti_dbinfos_train.pkl',
+    rate=1.0,
+    object_rot_range=[0.0, 0.0],
+    prepare=dict(
+        filter_by_difficulty=[-1],
+        filter_by_min_points=dict(
+            Car=5,
+            Pedestrian=5,
+            Cyclist=5,
+        )),
+    classes=class_names,
+    sample_groups=dict(
+        Car=15,
+        Pedestrian=10,
+        Cyclist=10,
+    ))
+
+train_pipeline = [
+    dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
+    dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
+    dict(type='ObjectSample', db_sampler=db_sampler),
+    dict(
+        type='ObjectNoise',
+        num_try=100,
+        loc_noise_std=[1.0, 1.0, 0.1],
+        global_rot_range=[0.0, 0.0],
+        rot_uniform_noise=[-0.78539816, 0.78539816]),
+    dict(type='RandomFlip3D', flip_ratio=0.5),
+    dict(
+        type='GlobalRotScale',
+        rot_uniform_noise=[-0.78539816, 0.78539816],
+        scaling_uniform_noise=[0.95, 1.05]),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='PointShuffle'),
+    dict(type='DefaultFormatBundle3D', class_names=class_names),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d']),
+]
+test_pipeline = [
+    dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(
+        type='DefaultFormatBundle3D',
+        class_names=class_names,
+        with_label=False),
+    dict(type='Collect3D', keys=['points']),
+]
+
+data = dict(
+    samples_per_gpu=6,
+    workers_per_gpu=4,
+    train=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_train.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=train_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=False),
+    val=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True),
+    test=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True))
+# optimizer
+lr = 0.001  # max learning rate
+optimizer = dict(
+    type='AdamW',
+    lr=lr,
+    betas=(0.95, 0.99),  # the momentum is change during training
+    weight_decay=0.01)
+optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))
+# learning policy
+lr_config = dict(
+    policy='cyclic',
+    target_ratio=(10, 1e-4),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+momentum_config = dict(
+    policy='cyclic',
+    target_ratio=(0.85 / 0.95, 1),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+checkpoint_config = dict(interval=1)
+evaluation = dict(interval=2)
+# yapf:disable
+log_config = dict(
+    interval=50,
+    hooks=[
+        dict(type='TextLoggerHook'),
+        dict(type='TensorboardLoggerHook')
+    ])
+# yapf:enable
+# runtime settings
+total_epochs = 80
+dist_params = dict(backend='nccl')
+log_level = 'INFO'
+work_dir = './work_dirs/pp_secfpn_80e'
+load_from = None
+resume_from = None
+workflow = [('train', 1)]

configs/benchmark/hv_second_secfpn_6x8_80e_pcdet_kitti-3d-3class.py

+# model settings
+voxel_size = [0.05, 0.05, 0.1]
+point_cloud_range = [0, -40, -3, 70.4, 40, 1]
+
+model = dict(
+    type='VoxelNet',
+    voxel_layer=dict(
+        max_num_points=5,
+        point_cloud_range=point_cloud_range,
+        voxel_size=voxel_size,
+        max_voxels=(16000, 40000)),
+    voxel_encoder=dict(type='HardSimpleVFE'),
+    middle_encoder=dict(
+        type='SparseEncoder',
+        in_channels=4,
+        sparse_shape=[41, 1600, 1408],
+        order=('conv', 'norm', 'act')),
+    backbone=dict(
+        type='SECOND',
+        in_channels=256,
+        layer_nums=[5, 5],
+        layer_strides=[1, 2],
+        out_channels=[128, 256]),
+    neck=dict(
+        type='SECONDFPN',
+        in_channels=[128, 256],
+        upsample_strides=[1, 2],
+        out_channels=[256, 256]),
+    bbox_head=dict(
+        type='Anchor3DHead',
+        num_classes=3,
+        in_channels=512,
+        feat_channels=512,
+        use_direction_classifier=True,
+        anchor_generator=dict(
+            type='Anchor3DRangeGenerator',
+            ranges=[
+                [0, -40.0, -0.6, 70.4, 40.0, -0.6],
+                [0, -40.0, -0.6, 70.4, 40.0, -0.6],
+                [0, -40.0, -1.78, 70.4, 40.0, -1.78],
+            ],
+            sizes=[[0.6, 0.8, 1.73], [0.6, 1.76, 1.73], [1.6, 3.9, 1.56]],
+            rotations=[0, 1.57],
+            reshape_out=False),
+        diff_rad_by_sin=True,
+        bbox_coder=dict(type='DeltaXYZWLHRBBoxCoder'),
+        loss_cls=dict(
+            type='FocalLoss',
+            use_sigmoid=True,
+            gamma=2.0,
+            alpha=0.25,
+            loss_weight=1.0),
+        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
+        loss_dir=dict(
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2)))
+# model training and testing settings
+train_cfg = dict(
+    assigner=[
+        dict(  # for Pedestrian
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.5,
+            neg_iou_thr=0.35,
+            min_pos_iou=0.35,
+            ignore_iof_thr=-1),
+        dict(  # for Cyclist
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.5,
+            neg_iou_thr=0.35,
+            min_pos_iou=0.35,
+            ignore_iof_thr=-1),
+        dict(  # for Car
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.6,
+            neg_iou_thr=0.45,
+            min_pos_iou=0.45,
+            ignore_iof_thr=-1),
+    ],
+    allowed_border=0,
+    pos_weight=-1,
+    debug=False)
+test_cfg = dict(
+    use_rotate_nms=True,
+    nms_across_levels=False,
+    nms_thr=0.01,
+    score_thr=0.1,
+    min_bbox_size=0,
+    nms_pre=100,
+    max_num=50)
+
+# dataset settings
+dataset_type = 'KittiDataset'
+data_root = 'data/kitti/'
+class_names = ['Pedestrian', 'Cyclist', 'Car']
+input_modality = dict(use_lidar=False, use_camera=False)
+db_sampler = dict(
+    data_root=data_root,
+    info_path=data_root + 'kitti_dbinfos_train.pkl',
+    rate=1.0,
+    object_rot_range=[0.0, 0.0],
+    prepare=dict(
+        filter_by_difficulty=[-1],
+        filter_by_min_points=dict(
+            Car=5,
+            Pedestrian=10,
+            Cyclist=10,
+        )),
+    classes=class_names,
+    sample_groups=dict(
+        Car=12,
+        Pedestrian=6,
+        Cyclist=6,
+    ))
+file_client_args = dict(backend='disk')
+# file_client_args = dict(
+#     backend='petrel', path_mapping=dict(data='s3://kitti_data/'))
+
+train_pipeline = [
+    dict(
+        type='LoadPointsFromFile',
+        load_dim=4,
+        use_dim=4,
+        file_client_args=file_client_args),
+    dict(
+        type='LoadAnnotations3D',
+        with_bbox_3d=True,
+        with_label_3d=True,
+        file_client_args=file_client_args),
+    dict(type='ObjectSample', db_sampler=db_sampler),
+    dict(
+        type='ObjectNoise',
+        num_try=100,
+        loc_noise_std=[1.0, 1.0, 0.1],
+        global_rot_range=[0.0, 0.0],
+        rot_uniform_noise=[-0.78539816, 0.78539816]),
+    dict(type='RandomFlip3D', flip_ratio=0.5),
+    dict(
+        type='GlobalRotScale',
+        rot_uniform_noise=[-0.78539816, 0.78539816],
+        scaling_uniform_noise=[0.95, 1.05]),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='PointShuffle'),
+    dict(type='DefaultFormatBundle3D', class_names=class_names),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
+]
+test_pipeline = [
+    dict(
+        type='LoadPointsFromFile',
+        load_dim=4,
+        use_dim=4,
+        file_client_args=file_client_args),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(
+        type='DefaultFormatBundle3D',
+        class_names=class_names,
+        with_label=False),
+    dict(type='Collect3D', keys=['points'])
+]
+
+data = dict(
+    samples_per_gpu=6,
+    workers_per_gpu=4,
+    train=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_train.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=train_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=False),
+    val=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True),
+    test=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True))
+# optimizer
+lr = 0.0018  # max learning rate
+optimizer = dict(type='AdamW', lr=lr, betas=(0.95, 0.99), weight_decay=0.01)
+optimizer_config = dict(grad_clip=dict(max_norm=10, norm_type=2))
+lr_config = dict(
+    policy='cyclic',
+    target_ratio=(10, 1e-4),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+momentum_config = dict(
+    policy='cyclic',
+    target_ratio=(0.85 / 0.95, 1),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+checkpoint_config = dict(interval=1)
+evaluation = dict(interval=2)
+# yapf:disable
+log_config = dict(
+    interval=50,
+    hooks=[
+        dict(type='TextLoggerHook'),
+        dict(type='TensorboardLoggerHook')
+    ])
+# yapf:enable
+# runtime settings
+total_epochs = 80
+dist_params = dict(backend='nccl')
+log_level = 'INFO'
+work_dir = './work_dirs/sec_secfpn_80e'
+load_from = None
+resume_from = None
+workflow = [('train', 1)]

configs/kitti/dv_mvx-v2_second_secfpn_fpn-fusion_adamw_2x8_80e_kitti-3d-3class.py

 # model settings
 voxel_size = [0.05, 0.05, 0.1]
-point_cloud_range = [0, -40, -3, 70.4, 40, 1]  # velodyne coordinates, x, y, z
+point_cloud_range = [0, -40, -3, 70.4, 40, 1]
 
 model = dict(
     type='DynamicMVXFasterRCNNV2',
@@ -21,10 +21,10 @@ model = dict(
         out_channels=256,
         num_outs=5),
     pts_voxel_layer=dict(
-        max_num_points=-1,  # max_points_per_voxel
+        max_num_points=-1,
         point_cloud_range=point_cloud_range,
         voxel_size=voxel_size,
-        max_voxels=(-1, -1),  # (training, testing) max_coxels
+        max_voxels=(-1, -1),
     ),
     pts_voxel_encoder=dict(
         type='DynamicVFE',
@@ -44,8 +44,7 @@ model = dict(
             img_levels=[0, 1, 2, 3, 4],
             align_corners=False,
             activate_out=True,
-            fuse_out=False),
-    ),
+            fuse_out=False)),
     pts_middle_encoder=dict(
         type='SparseEncoder',
         in_channels=128,
@@ -56,14 +55,12 @@ model = dict(
         in_channels=256,
         layer_nums=[5, 5],
         layer_strides=[1, 2],
-        out_channels=[128, 256],
-    ),
+        out_channels=[128, 256]),
     pts_neck=dict(
         type='SECONDFPN',
         in_channels=[128, 256],
         upsample_strides=[1, 2],
-        out_channels=[256, 256],
-    ),
+        out_channels=[256, 256]),
     pts_bbox_head=dict(
         type='Anchor3DHead',
         num_classes=3,
@@ -92,8 +89,7 @@ model = dict(
             loss_weight=1.0),
         loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
         loss_dir=dict(
-            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2),
-    ))
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2)))
 # model training and testing settings
 train_cfg = dict(
     pts=dict(
@@ -128,7 +124,7 @@ test_cfg = dict(
         use_rotate_nms=True,
         nms_across_levels=False,
         nms_thr=0.01,
-        score_thr=0.3,
+        score_thr=0.1,
         min_bbox_size=0,
         nms_pre=100,
         max_num=50))
@@ -139,12 +135,7 @@ data_root = 'data/kitti/'
 class_names = ['Pedestrian', 'Cyclist', 'Car']
 img_norm_cfg = dict(
     mean=[103.530, 116.280, 123.675], std=[1.0, 1.0, 1.0], to_rgb=False)
-input_modality = dict(
-    use_lidar=True,
-    use_depth=False,
-    use_lidar_intensity=True,
-    use_camera=True,
-)
+input_modality = dict(use_lidar=True, use_camera=True)
 db_sampler = dict(
     type='MMDataBaseSampler',
     data_root=data_root,
@@ -157,17 +148,8 @@ db_sampler = dict(
     collision_thr=[0, 0.3, 0.5, 0.7],
     prepare=dict(
         filter_by_difficulty=[-1],
-        filter_by_min_points=dict(
-            Car=5,
-            Pedestrian=10,
-            Cyclist=10,
-        ),
-    ),
-    sample_groups=dict(
-        Car=12,
-        Pedestrian=6,
-        Cyclist=6,
-    ),
+        filter_by_min_points=dict(Car=5, Pedestrian=10, Cyclist=10)),
+    sample_groups=dict(Car=12, Pedestrian=6, Cyclist=6),
     classes=class_names)
 train_pipeline = [
     dict(
@@ -189,10 +171,7 @@ train_pipeline = [
     dict(type='DefaultFormatBundle3D', class_names=class_names),
     dict(
         type='Collect3D',
-        keys=[
-            'points', 'img', 'gt_bboxes_3d', 'gt_bboxes', 'gt_labels',
-            'gt_labels_3d'
-        ]),
+        keys=['points', 'img', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
     dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
@@ -214,7 +193,7 @@ test_pipeline = [
         type='DefaultFormatBundle3D',
         class_names=class_names,
         with_label=False),
-    dict(type='Collect3D', keys=['points', 'img']),
+    dict(type='Collect3D', keys=['points', 'img'])
 ]
 
 data = dict(

configs/kitti/dv_pointpillars_secfpn_6x8_160e_kitti-3d-car.py

@@ -5,37 +5,30 @@ point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
 model = dict(
     type='DynamicVoxelNet',
     voxel_layer=dict(
-        max_num_points=-1,  # set -1 for dynamic voxel
-        point_cloud_range=point_cloud_range,  # velodyne coordinates, x, y, z
+        max_num_points=-1,
+        point_cloud_range=point_cloud_range,
         voxel_size=voxel_size,
-        max_voxels=(-1, -1),  # set -1 for dynamic voxel
-    ),
+        max_voxels=(-1, -1)),
     voxel_encoder=dict(
         type='DynamicPillarFeatureNet',
         in_channels=4,
         feat_channels=[64],
         with_distance=False,
         voxel_size=voxel_size,
-        point_cloud_range=point_cloud_range,
-    ),
+        point_cloud_range=point_cloud_range),
     middle_encoder=dict(
-        type='PointPillarsScatter',
-        in_channels=64,
-        output_shape=[496, 432],
-    ),
+        type='PointPillarsScatter', in_channels=64, output_shape=[496, 432]),
     backbone=dict(
         type='SECOND',
         in_channels=64,
         layer_nums=[3, 5, 5],
         layer_strides=[2, 2, 2],
-        out_channels=[64, 128, 256],
-    ),
+        out_channels=[64, 128, 256]),
     neck=dict(
         type='SECONDFPN',
         in_channels=[64, 128, 256],
         upsample_strides=[1, 2, 4],
-        out_channels=[128, 128, 128],
-    ),
+        out_channels=[128, 128, 128]),
     bbox_head=dict(
         type='Anchor3DHead',
         num_classes=1,
@@ -58,9 +51,7 @@ model = dict(
             loss_weight=1.0),
         loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
         loss_dir=dict(
-            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2),
-    ),
-)
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2)))
 # model training and testing settings
 train_cfg = dict(
     assigner=dict(
@@ -77,7 +68,7 @@ test_cfg = dict(
     use_rotate_nms=True,
     nms_across_levels=False,
     nms_thr=0.01,
-    score_thr=0.3,
+    score_thr=0.1,
     min_bbox_size=0,
     nms_pre=100,
     max_num=50)
@@ -86,15 +77,7 @@ test_cfg = dict(
 dataset_type = 'KittiDataset'
 data_root = 'data/kitti/'
 class_names = ['Car']
-img_norm_cfg = dict(
-    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
-input_modality = dict(
-    use_lidar=False,
-    use_lidar_reduced=True,
-    use_depth=False,
-    use_lidar_intensity=True,
-    use_camera=False,
-)
+input_modality = dict(use_lidar=True, use_camera=False)
 db_sampler = dict(
     root_path=data_root,
     info_path=data_root + 'kitti_dbinfos_train.pkl',
@@ -126,7 +109,7 @@ train_pipeline = [
     dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='PointShuffle'),
     dict(type='DefaultFormatBundle3D', class_names=class_names),
-    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d']),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
     dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
@@ -135,7 +118,7 @@ test_pipeline = [
         type='DefaultFormatBundle3D',
         class_names=class_names,
         with_label=False),
-    dict(type='Collect3D', keys=['points']),
+    dict(type='Collect3D', keys=['points'])
 ]
 
 data = dict(
@@ -180,14 +163,12 @@ lr_config = dict(
     policy='cyclic',
     target_ratio=(10, 1e-4),
     cyclic_times=1,
-    step_ratio_up=0.4,
-)
+    step_ratio_up=0.4)
 momentum_config = dict(
     policy='cyclic',
     target_ratio=(0.85 / 0.95, 1),
     cyclic_times=1,
-    step_ratio_up=0.4,
-)
+    step_ratio_up=0.4)
 checkpoint_config = dict(interval=1)
 evaluation = dict(interval=2)
 # yapf:disable

configs/kitti/dv_second_secfpn_2x8_cosine_80e_kitti-3d-3class.py

 # model settings
 voxel_size = [0.05, 0.05, 0.1]
-point_cloud_range = [0, -40, -3, 70.4, 40, 1]  # velodyne coordinates, x, y, z
+point_cloud_range = [0, -40, -3, 70.4, 40, 1]
 
 model = dict(
     type='DynamicVoxelNet',
@@ -8,7 +8,7 @@ model = dict(
         max_num_points=-1,  # max_points_per_voxel
         point_cloud_range=point_cloud_range,
         voxel_size=voxel_size,
-        max_voxels=(-1, -1),  # (training, testing) max_coxels
+        max_voxels=(-1, -1)  # (training, testing) max_coxels
     ),
     voxel_encoder=dict(
         type='DynamicSimpleVFE',
@@ -24,14 +24,12 @@ model = dict(
         in_channels=256,
         layer_nums=[5, 5],
         layer_strides=[1, 2],
-        out_channels=[128, 256],
-    ),
+        out_channels=[128, 256]),
     neck=dict(
         type='SECONDFPN',
         in_channels=[128, 256],
         upsample_strides=[1, 2],
-        out_channels=[256, 256],
-    ),
+        out_channels=[256, 256]),
     bbox_head=dict(
         type='Anchor3DHead',
         num_classes=3,
@@ -60,9 +58,7 @@ model = dict(
             loss_weight=1.0),
         loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
         loss_dir=dict(
-            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2),
-    ),
-)
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2)))
 # model training and testing settings
 train_cfg = dict(
     assigner=[
@@ -95,7 +91,7 @@ test_cfg = dict(
     use_rotate_nms=True,
     nms_across_levels=False,
     nms_thr=0.01,
-    score_thr=0.3,
+    score_thr=0.1,
     min_bbox_size=0,
     nms_pre=100,
     max_num=50)
@@ -104,15 +100,7 @@ test_cfg = dict(
 dataset_type = 'KittiDataset'
 data_root = 'data/kitti/'
 class_names = ['Pedestrian', 'Cyclist', 'Car']
-img_norm_cfg = dict(
-    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
-input_modality = dict(
-    use_lidar=False,
-    use_lidar_reduced=True,
-    use_depth=False,
-    use_lidar_intensity=True,
-    use_camera=True,
-)
+input_modality = dict(use_lidar=True, use_camera=False)
 db_sampler = dict(
     data_root=data_root,
     info_path=data_root + 'kitti_dbinfos_train.pkl',
@@ -120,17 +108,8 @@ db_sampler = dict(
     object_rot_range=[0.0, 0.0],
     prepare=dict(
         filter_by_difficulty=[-1],
-        filter_by_min_points=dict(
-            Car=5,
-            Pedestrian=10,
-            Cyclist=10,
-        ),
-    ),
-    sample_groups=dict(
-        Car=12,
-        Pedestrian=6,
-        Cyclist=6,
-    ),
+        filter_by_min_points=dict(Car=5, Pedestrian=10, Cyclist=10)),
+    sample_groups=dict(Car=12, Pedestrian=6, Cyclist=6),
     classes=class_names)
 train_pipeline = [
     dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
@@ -152,7 +131,7 @@ train_pipeline = [
     dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='PointShuffle'),
     dict(type='DefaultFormatBundle3D', class_names=class_names),
-    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d']),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
     dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
@@ -161,22 +140,25 @@ test_pipeline = [
         type='DefaultFormatBundle3D',
         class_names=class_names,
         with_label=False),
-    dict(type='Collect3D', keys=['points']),
+    dict(type='Collect3D', keys=['points'])
 ]
 
 data = dict(
     samples_per_gpu=2,
     workers_per_gpu=2,
     train=dict(
-        type=dataset_type,
-        data_root=data_root,
-        ann_file=data_root + 'kitti_infos_train.pkl',
-        split='training',
-        pts_prefix='velodyne_reduced',
-        pipeline=train_pipeline,
-        modality=input_modality,
-        classes=class_names,
-        test_mode=False),
+        type='RepeatDataset',
+        times=2,
+        dataset=dict(
+            type=dataset_type,
+            data_root=data_root,
+            ann_file=data_root + 'kitti_infos_train.pkl',
+            split='training',
+            pts_prefix='velodyne_reduced',
+            pipeline=train_pipeline,
+            modality=input_modality,
+            classes=class_names,
+            test_mode=False)),
     val=dict(
         type=dataset_type,
         data_root=data_root,
@@ -213,7 +195,7 @@ lr_config = dict(
     min_lr_ratio=1e-5)
 momentum_config = None
 checkpoint_config = dict(interval=1)
-evaluation = dict(interval=2)
+evaluation = dict(interval=1)
 # yapf:disable
 log_config = dict(
     interval=50,
@@ -223,7 +205,7 @@ log_config = dict(
     ])
 # yapf:enable
 # runtime settings
-total_epochs = 80
+total_epochs = 40
 dist_params = dict(backend='nccl', port=29502)
 log_level = 'INFO'
 work_dir = './work_dirs/sec_secfpn_80e'

configs/kitti/dv_second_secfpn_6x8_80e_kitti-3d-car.py

@@ -8,7 +8,7 @@ model = dict(
         max_num_points=-1,  # max_points_per_voxel
         point_cloud_range=point_cloud_range,
         voxel_size=voxel_size,
-        max_voxels=(-1, -1),  # (training, testing) max_coxels
+        max_voxels=(-1, -1)  # (training, testing) max_coxels
     ),
     voxel_encoder=dict(
         type='DynamicSimpleVFE',
@@ -24,14 +24,12 @@ model = dict(
         in_channels=256,
         layer_nums=[5, 5],
         layer_strides=[1, 2],
-        out_channels=[128, 256],
-    ),
+        out_channels=[128, 256]),
     neck=dict(
         type='SECONDFPN',
         in_channels=[128, 256],
         upsample_strides=[1, 2],
-        out_channels=[256, 256],
-    ),
+        out_channels=[256, 256]),
     bbox_head=dict(
         type='Anchor3DHead',
         num_classes=1,
@@ -54,9 +52,7 @@ model = dict(
             loss_weight=1.0),
         loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
         loss_dir=dict(
-            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2),
-    ),
-)
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2)))
 # model training and testing settings
 train_cfg = dict(
     assigner=dict(
@@ -73,7 +69,7 @@ test_cfg = dict(
     use_rotate_nms=True,
     nms_across_levels=False,
     nms_thr=0.01,
-    score_thr=0.3,
+    score_thr=0.1,
     min_bbox_size=0,
     nms_pre=100,
     max_num=50)
@@ -82,15 +78,7 @@ test_cfg = dict(
 dataset_type = 'KittiDataset'
 data_root = 'data/kitti/'
 class_names = ['Car']
-img_norm_cfg = dict(
-    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
-input_modality = dict(
-    use_lidar=False,
-    use_lidar_reduced=True,
-    use_depth=False,
-    use_lidar_intensity=True,
-    use_camera=True,
-)
+input_modality = dict(use_lidar=True, use_camera=False)
 db_sampler = dict(
     data_root=data_root,
     info_path=data_root + 'kitti_dbinfos_train.pkl',
@@ -137,15 +125,18 @@ data = dict(
     samples_per_gpu=6,
     workers_per_gpu=4,
     train=dict(
-        type=dataset_type,
-        data_root=data_root,
-        ann_file=data_root + 'kitti_infos_train.pkl',
-        split='training',
-        pts_prefix='velodyne_reduced',
-        pipeline=train_pipeline,
-        modality=input_modality,
-        classes=class_names,
-        test_mode=False),
+        type='RepeatDataset',
+        times=2,
+        dataset=dict(
+            type=dataset_type,
+            data_root=data_root,
+            ann_file=data_root + 'kitti_infos_train.pkl',
+            split='training',
+            pts_prefix='velodyne_reduced',
+            pipeline=train_pipeline,
+            modality=input_modality,
+            classes=class_names,
+            test_mode=False)),
     val=dict(
         type=dataset_type,
         data_root=data_root,
@@ -183,7 +174,7 @@ momentum_config = dict(
     step_ratio_up=0.4,
 )
 checkpoint_config = dict(interval=1)
-evaluation = dict(interval=2)
+evaluation = dict(interval=1)
 # yapf:disable
 log_config = dict(
     interval=50,
@@ -193,7 +184,7 @@ log_config = dict(
     ])
 # yapf:enable
 # runtime settings
-total_epochs = 80
+total_epochs = 40
 dist_params = dict(backend='nccl')
 log_level = 'INFO'
 work_dir = './work_dirs/sec_secfpn_80e'

configs/kitti/hv_PartA2_secfpn_4x8_cyclic_80e_kitti-3d-3class.py → configs/kitti/hv_PartA2_secfpn_2x8_cyclic_80e_kitti-3d-3class.py

@@ -198,12 +198,7 @@ test_cfg = dict(
 dataset_type = 'KittiDataset'
 data_root = 'data/kitti/'
 class_names = ['Pedestrian', 'Cyclist', 'Car']
-input_modality = dict(
-    use_lidar=False,
-    use_lidar_reduced=True,
-    use_depth=False,
-    use_lidar_intensity=True,
-    use_camera=False)
+input_modality = dict(use_lidar=True, use_camera=False)
 db_sampler = dict(
     data_root=data_root,
     info_path=data_root + 'kitti_dbinfos_train.pkl',
@@ -211,17 +206,9 @@ db_sampler = dict(
     object_rot_range=[0.0, 0.0],
     prepare=dict(
         filter_by_difficulty=[-1],
-        filter_by_min_points=dict(
-            Car=5,
-            Pedestrian=10,
-            Cyclist=10,
-        )),
+        filter_by_min_points=dict(Car=5, Pedestrian=10, Cyclist=10)),
     classes=class_names,
-    sample_groups=dict(
-        Car=12,
-        Pedestrian=6,
-        Cyclist=6,
-    ))
+    sample_groups=dict(Car=12, Pedestrian=6, Cyclist=6))
 train_pipeline = [
     dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
     dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
@@ -258,15 +245,18 @@ data = dict(
     samples_per_gpu=2,
     workers_per_gpu=2,
     train=dict(
-        type=dataset_type,
-        data_root=data_root,
-        ann_file=data_root + 'kitti_infos_train.pkl',
-        split='training',
-        pts_prefix='velodyne_reduced',
-        pipeline=train_pipeline,
-        modality=input_modality,
-        classes=class_names,
-        test_mode=False),
+        type='RepeatDataset',
+        times=2,
+        dataset=dict(
+            type=dataset_type,
+            data_root=data_root,
+            ann_file=data_root + 'kitti_infos_train.pkl',
+            split='training',
+            pts_prefix='velodyne_reduced',
+            pipeline=train_pipeline,
+            modality=input_modality,
+            classes=class_names,
+            test_mode=False)),
     val=dict(
         type=dataset_type,
         data_root=data_root,
@@ -302,7 +292,7 @@ momentum_config = dict(
     cyclic_times=1,
     step_ratio_up=0.4)
 checkpoint_config = dict(interval=1)
-evaluation = dict(interval=2)
+evaluation = dict(interval=1)
 # yapf:disable
 log_config = dict(
     interval=50,
@@ -312,8 +302,8 @@ log_config = dict(
     ])
 # yapf:enable
 # runtime settings
-total_epochs = 80
-dist_params = dict(backend='nccl')
+total_epochs = 40
+dist_params = dict(backend='nccl', port=29506)
 log_level = 'INFO'
 find_unused_parameters = True
 work_dir = './work_dirs/parta2_secfpn_80e'

configs/kitti/hv_PartA2_secfpn_4x8_cyclic_80e_kitti-3d-car.py → configs/kitti/hv_PartA2_secfpn_2x8_cyclic_80e_kitti-3d-car.py

@@ -164,21 +164,13 @@ test_cfg = dict(
 dataset_type = 'KittiDataset'
 data_root = 'data/kitti/'
 class_names = ['Car']
-input_modality = dict(
-    use_lidar=False,
-    use_lidar_reduced=True,
-    use_depth=False,
-    use_lidar_intensity=True,
-    use_camera=False)
+input_modality = dict(use_lidar=True, use_camera=False)
 db_sampler = dict(
     data_root=data_root,
     info_path=data_root + 'kitti_dbinfos_train.pkl',
     rate=1.0,
     object_rot_range=[0.0, 0.0],
-    prepare=dict(
-        filter_by_difficulty=[-1],
-        filter_by_min_points=dict(Car=5),
-    ),
+    prepare=dict(filter_by_difficulty=[-1], filter_by_min_points=dict(Car=5)),
     classes=class_names,
     sample_groups=dict(Car=15))
 train_pipeline = [
@@ -217,15 +209,18 @@ data = dict(
     samples_per_gpu=2,
     workers_per_gpu=2,
     train=dict(
-        type=dataset_type,
-        data_root=data_root,
-        ann_file=data_root + 'kitti_infos_train.pkl',
-        split='training',
-        pts_prefix='velodyne_reduced',
-        pipeline=train_pipeline,
-        modality=input_modality,
-        classes=class_names,
-        test_mode=False),
+        type='RepeatDataset',
+        times=2,
+        dataset=dict(
+            type=dataset_type,
+            data_root=data_root,
+            ann_file=data_root + 'kitti_infos_train.pkl',
+            split='training',
+            pts_prefix='velodyne_reduced',
+            pipeline=train_pipeline,
+            modality=input_modality,
+            classes=class_names,
+            test_mode=False)),
     val=dict(
         type=dataset_type,
         data_root=data_root,
@@ -261,7 +256,7 @@ momentum_config = dict(
     cyclic_times=1,
     step_ratio_up=0.4)
 checkpoint_config = dict(interval=1)
-evaluation = dict(interval=2)
+evaluation = dict(interval=1)
 # yapf:disable
 log_config = dict(
     interval=50,
@@ -271,7 +266,7 @@ log_config = dict(
     ])
 # yapf:enable
 # runtime settings
-total_epochs = 80
+total_epochs = 40
 dist_params = dict(backend='nccl')
 log_level = 'INFO'
 find_unused_parameters = True

configs/kitti/hv_pointpillars_secfpn_6x8_160e_kitti-3d-3class.py

+# model settings
+point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
+voxel_size = [0.16, 0.16, 4]
+model = dict(
+    type='VoxelNet',
+    voxel_layer=dict(
+        max_num_points=32,
+        point_cloud_range=point_cloud_range,
+        voxel_size=voxel_size,
+        max_voxels=(16000, 40000)),
+    voxel_encoder=dict(
+        type='PillarFeatureNet',
+        in_channels=4,
+        feat_channels=[64],
+        with_distance=False,
+        voxel_size=voxel_size,
+        point_cloud_range=point_cloud_range),
+    middle_encoder=dict(
+        type='PointPillarsScatter', in_channels=64, output_shape=[496, 432]),
+    backbone=dict(
+        type='SECOND',
+        in_channels=64,
+        layer_nums=[3, 5, 5],
+        layer_strides=[2, 2, 2],
+        out_channels=[64, 128, 256]),
+    neck=dict(
+        type='SECONDFPN',
+        in_channels=[64, 128, 256],
+        upsample_strides=[1, 2, 4],
+        out_channels=[128, 128, 128]),
+    bbox_head=dict(
+        type='Anchor3DHead',
+        num_classes=3,
+        in_channels=384,
+        feat_channels=384,
+        use_direction_classifier=True,
+        anchor_generator=dict(
+            type='Anchor3DRangeGenerator',
+            ranges=[
+                [0, -39.68, -0.6, 70.4, 39.68, -0.6],
+                [0, -39.68, -0.6, 70.4, 39.68, -0.6],
+                [0, -39.68, -1.78, 70.4, 39.68, -1.78],
+            ],
+            sizes=[[0.6, 0.8, 1.73], [0.6, 1.76, 1.73], [1.6, 3.9, 1.56]],
+            rotations=[0, 1.57],
+            reshape_out=False),
+        diff_rad_by_sin=True,
+        bbox_coder=dict(type='DeltaXYZWLHRBBoxCoder'),
+        loss_cls=dict(
+            type='FocalLoss',
+            use_sigmoid=True,
+            gamma=2.0,
+            alpha=0.25,
+            loss_weight=1.0),
+        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
+        loss_dir=dict(
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2)))
+# model training and testing settings
+train_cfg = dict(
+    assigner=[
+        dict(  # for Pedestrian
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.5,
+            neg_iou_thr=0.35,
+            min_pos_iou=0.35,
+            ignore_iof_thr=-1),
+        dict(  # for Cyclist
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.5,
+            neg_iou_thr=0.35,
+            min_pos_iou=0.35,
+            ignore_iof_thr=-1),
+        dict(  # for Car
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.6,
+            neg_iou_thr=0.45,
+            min_pos_iou=0.45,
+            ignore_iof_thr=-1),
+    ],
+    allowed_border=0,
+    pos_weight=-1,
+    debug=False)
+test_cfg = dict(
+    use_rotate_nms=True,
+    nms_across_levels=False,
+    nms_thr=0.01,
+    score_thr=0.1,
+    min_bbox_size=0,
+    nms_pre=100,
+    max_num=50)
+
+# dataset settings
+dataset_type = 'KittiDataset'
+data_root = 'data/kitti/'
+class_names = ['Pedestrian', 'Cyclist', 'Car']
+input_modality = dict(use_lidar=True, use_camera=False)
+db_sampler = dict(
+    data_root=data_root,
+    info_path=data_root + 'kitti_dbinfos_train.pkl',
+    rate=1.0,
+    object_rot_range=[0.0, 0.0],
+    prepare=dict(
+        filter_by_difficulty=[-1],
+        filter_by_min_points=dict(
+            Car=5,
+            Pedestrian=10,
+            Cyclist=10,
+        )),
+    classes=class_names,
+    sample_groups=dict(
+        Car=12,
+        Pedestrian=6,
+        Cyclist=6,
+    ))
+
+train_pipeline = [
+    dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
+    dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
+    dict(type='ObjectSample', db_sampler=db_sampler),
+    dict(
+        type='ObjectNoise',
+        num_try=100,
+        loc_noise_std=[0.25, 0.25, 0.25],
+        global_rot_range=[0.0, 0.0],
+        rot_uniform_noise=[-0.15707963267, 0.15707963267]),
+    dict(type='RandomFlip3D', flip_ratio=0.5),
+    dict(
+        type='GlobalRotScale',
+        rot_uniform_noise=[-0.78539816, 0.78539816],
+        scaling_uniform_noise=[0.95, 1.05]),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='PointShuffle'),
+    dict(type='DefaultFormatBundle3D', class_names=class_names),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
+]
+test_pipeline = [
+    dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(
+        type='DefaultFormatBundle3D',
+        class_names=class_names,
+        with_label=False),
+    dict(type='Collect3D', keys=['points'])
+]
+
+data = dict(
+    samples_per_gpu=6,
+    workers_per_gpu=4,
+    train=dict(
+        type='RepeatDataset',
+        times=2,
+        dataset=dict(
+            type=dataset_type,
+            data_root=data_root,
+            ann_file=data_root + 'kitti_infos_train.pkl',
+            split='training',
+            pts_prefix='velodyne_reduced',
+            pipeline=train_pipeline,
+            modality=input_modality,
+            classes=class_names,
+            test_mode=False)),
+    val=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True),
+    test=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True))
+# optimizer
+lr = 0.001  # max learning rate
+optimizer = dict(
+    type='AdamW',
+    lr=lr,
+    betas=(0.95, 0.99),  # the momentum is change during training
+    weight_decay=0.01)
+optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))
+# learning policy
+lr_config = dict(
+    policy='cyclic',
+    target_ratio=(10, 1e-4),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+momentum_config = dict(
+    policy='cyclic',
+    target_ratio=(0.85 / 0.95, 1),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+checkpoint_config = dict(interval=1)
+evaluation = dict(interval=2)
+# yapf:disable
+log_config = dict(
+    interval=50,
+    hooks=[
+        dict(type='TextLoggerHook'),
+        dict(type='TensorboardLoggerHook')
+    ])
+# yapf:enable
+# runtime settings
+total_epochs = 80
+dist_params = dict(backend='nccl')
+log_level = 'INFO'
+work_dir = './work_dirs/pp_secfpn_80e'
+load_from = None
+resume_from = None
+workflow = [('train', 1)]

configs/kitti/hv_pointpillars_secfpn_6x8_160e_kitti-3d-car.py

 # model settings
+voxel_size = [0.16, 0.16, 4]
 point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
 model = dict(
     type='VoxelNet',
     voxel_layer=dict(
-        max_num_points=64,  # max_points_per_voxel
-        point_cloud_range=point_cloud_range,  # velodyne coordinates, x, y, z
-        voxel_size=[0.16, 0.16, 4],
-        max_voxels=(12000, 20000),  # (training, testing) max_coxels
+        max_num_points=64,
+        point_cloud_range=point_cloud_range,
+        voxel_size=voxel_size,
+        max_voxels=(12000, 20000)  # (training, testing) max_coxels
     ),
     voxel_encoder=dict(
         type='PillarFeatureNet',
         in_channels=4,
         feat_channels=[64],
         with_distance=False,
-        # these two arguments should be consistent with the voxel_generator
-        voxel_size=[0.16, 0.16, 4],
+        voxel_size=voxel_size,
         point_cloud_range=point_cloud_range,
     ),
     middle_encoder=dict(
@@ -76,7 +76,7 @@ test_cfg = dict(
     use_rotate_nms=True,
     nms_across_levels=False,
     nms_thr=0.01,
-    score_thr=0.3,
+    score_thr=0.1,
     min_bbox_size=0,
     nms_pre=100,
     max_num=50)
@@ -85,15 +85,7 @@ test_cfg = dict(
 dataset_type = 'KittiDataset'
 data_root = 'data/kitti/'
 class_names = ['Car']
-img_norm_cfg = dict(
-    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
-input_modality = dict(
-    use_lidar=False,
-    use_lidar_reduced=True,
-    use_depth=False,
-    use_lidar_intensity=True,
-    use_camera=False,
-)
+input_modality = dict(use_lidar=True, use_camera=False)
 db_sampler = dict(
     data_root=data_root,
     info_path=data_root + 'kitti_dbinfos_train.pkl',
@@ -141,15 +133,18 @@ data = dict(
     samples_per_gpu=6,
     workers_per_gpu=4,
     train=dict(
-        type=dataset_type,
-        data_root=data_root,
-        ann_file=data_root + 'kitti_infos_train.pkl',
-        split='training',
-        pts_prefix='velodyne_reduced',
-        pipeline=train_pipeline,
-        modality=input_modality,
-        classes=class_names,
-        test_mode=False),
+        type='RepeatDataset',
+        times=2,
+        dataset=dict(
+            type=dataset_type,
+            data_root=data_root,
+            ann_file=data_root + 'kitti_infos_train.pkl',
+            split='training',
+            pts_prefix='velodyne_reduced',
+            pipeline=train_pipeline,
+            modality=input_modality,
+            classes=class_names,
+            test_mode=False)),
     val=dict(
         type=dataset_type,
         data_root=data_root,
@@ -192,7 +187,7 @@ momentum_config = dict(
     step_ratio_up=0.4,
 )
 checkpoint_config = dict(interval=1)
-evaluation = dict(interval=2)
+evaluation = dict(interval=1)
 # yapf:disable
 log_config = dict(
     interval=50,
@@ -202,7 +197,7 @@ log_config = dict(
     ])
 # yapf:enable
 # runtime settings
-total_epochs = 160
+total_epochs = 80
 dist_params = dict(backend='nccl')
 log_level = 'INFO'
 work_dir = './work_dirs/pp_secfpn_80e'

configs/kitti/hv_pointpillars_secfpn_6x8_80e_kitti-3d-car.py

+# model settings
+voxel_size = [0.16, 0.16, 4]
+point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
+model = dict(
+    type='VoxelNet',
+    voxel_layer=dict(
+        max_num_points=64,
+        point_cloud_range=point_cloud_range,
+        voxel_size=voxel_size,
+        max_voxels=(12000, 20000)  # (training, testing) max_coxels
+    ),
+    voxel_encoder=dict(
+        type='PillarFeatureNet',
+        in_channels=4,
+        feat_channels=[64],
+        with_distance=False,
+        voxel_size=voxel_size,
+        point_cloud_range=point_cloud_range,
+    ),
+    middle_encoder=dict(
+        type='PointPillarsScatter',
+        in_channels=64,
+        output_shape=[496, 432],
+    ),
+    backbone=dict(
+        type='SECOND',
+        in_channels=64,
+        layer_nums=[3, 5, 5],
+        layer_strides=[2, 2, 2],
+        out_channels=[64, 128, 256],
+    ),
+    neck=dict(
+        type='SECONDFPN',
+        in_channels=[64, 128, 256],
+        upsample_strides=[1, 2, 4],
+        out_channels=[128, 128, 128],
+    ),
+    bbox_head=dict(
+        type='Anchor3DHead',
+        num_classes=1,
+        in_channels=384,
+        feat_channels=384,
+        use_direction_classifier=True,
+        anchor_generator=dict(
+            type='Anchor3DRangeGenerator',
+            ranges=[[0, -39.68, -1.78, 69.12, 39.68, -1.78]],
+            sizes=[[1.6, 3.9, 1.56]],
+            rotations=[0, 1.57],
+            reshape_out=True),
+        diff_rad_by_sin=True,
+        bbox_coder=dict(type='DeltaXYZWLHRBBoxCoder'),
+        loss_cls=dict(
+            type='FocalLoss',
+            use_sigmoid=True,
+            gamma=2.0,
+            alpha=0.25,
+            loss_weight=1.0),
+        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
+        loss_dir=dict(
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2),
+    ),
+)
+# model training and testing settings
+train_cfg = dict(
+    assigner=dict(
+        type='MaxIoUAssigner',
+        iou_calculator=dict(type='BboxOverlapsNearest3D'),
+        pos_iou_thr=0.6,
+        neg_iou_thr=0.45,
+        min_pos_iou=0.45,
+        ignore_iof_thr=-1),
+    allowed_border=0,
+    pos_weight=-1,
+    debug=False)
+test_cfg = dict(
+    use_rotate_nms=True,
+    nms_across_levels=False,
+    nms_thr=0.01,
+    score_thr=0.1,
+    min_bbox_size=0,
+    nms_pre=100,
+    max_num=50)
+
+# dataset settings
+dataset_type = 'KittiDataset'
+data_root = 'data/kitti/'
+class_names = ['Car']
+input_modality = dict(use_lidar=True, use_camera=False)
+db_sampler = dict(
+    data_root=data_root,
+    info_path=data_root + 'kitti_dbinfos_train.pkl',
+    rate=1.0,
+    object_rot_range=[0.0, 0.0],
+    prepare=dict(
+        filter_by_difficulty=[-1],
+        filter_by_min_points=dict(Car=5),
+    ),
+    sample_groups=dict(Car=15),
+    classes=class_names)
+
+train_pipeline = [
+    dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
+    dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
+    dict(type='ObjectSample', db_sampler=db_sampler),
+    dict(
+        type='ObjectNoise',
+        num_try=100,
+        loc_noise_std=[0.25, 0.25, 0.25],
+        global_rot_range=[0.0, 0.0],
+        rot_uniform_noise=[-0.15707963267, 0.15707963267]),
+    dict(type='RandomFlip3D', flip_ratio=0.5),
+    dict(
+        type='GlobalRotScale',
+        rot_uniform_noise=[-0.78539816, 0.78539816],
+        scaling_uniform_noise=[0.95, 1.05]),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='PointShuffle'),
+    dict(type='DefaultFormatBundle3D', class_names=class_names),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d']),
+]
+test_pipeline = [
+    dict(type='LoadPointsFromFile', load_dim=4, use_dim=4),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(
+        type='DefaultFormatBundle3D',
+        class_names=class_names,
+        with_label=False),
+    dict(type='Collect3D', keys=['points']),
+]
+
+data = dict(
+    samples_per_gpu=6,
+    workers_per_gpu=4,
+    train=dict(
+        type='RepeatDataset',
+        times=2,
+        dataset=dict(
+            type=dataset_type,
+            data_root=data_root,
+            ann_file=data_root + 'kitti_infos_train.pkl',
+            split='training',
+            pts_prefix='velodyne_reduced',
+            pipeline=train_pipeline,
+            modality=input_modality,
+            classes=class_names,
+            test_mode=False)),
+    val=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True),
+    test=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True))
+# optimizer
+lr = 0.001  # max learning rate
+optimizer = dict(
+    type='AdamW',
+    lr=lr,
+    betas=(0.95, 0.99),  # the momentum is change during training
+    weight_decay=0.01)
+optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))
+# learning policy
+lr_config = dict(
+    policy='cyclic',
+    target_ratio=(10, 1e-4),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+momentum_config = dict(
+    policy='cyclic',
+    target_ratio=(0.85 / 0.95, 1),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+checkpoint_config = dict(interval=1)
+evaluation = dict(interval=1)
+# yapf:disable
+log_config = dict(
+    interval=50,
+    hooks=[
+        dict(type='TextLoggerHook'),
+        dict(type='TensorboardLoggerHook')
+    ])
+# yapf:enable
+# runtime settings
+total_epochs = 40
+dist_params = dict(backend='nccl')
+log_level = 'INFO'
+work_dir = './work_dirs/pp_secfpn_80e'
+load_from = None
+resume_from = None
+workflow = [('train', 1)]

configs/kitti/hv_second_secfpn_6x8_80e_kitti-3d-3class.py

+# model settings
+voxel_size = [0.05, 0.05, 0.1]
+point_cloud_range = [0, -40, -3, 70.4, 40, 1]
+
+model = dict(
+    type='VoxelNet',
+    voxel_layer=dict(
+        max_num_points=5,
+        point_cloud_range=point_cloud_range,
+        voxel_size=voxel_size,
+        max_voxels=(16000, 40000),  # (training, testing) max_coxels
+    ),
+    voxel_encoder=dict(type='HardSimpleVFE'),
+    middle_encoder=dict(
+        type='SparseEncoder',
+        in_channels=4,
+        sparse_shape=[41, 1600, 1408],
+        order=('conv', 'norm', 'act')),
+    backbone=dict(
+        type='SECOND',
+        in_channels=256,
+        layer_nums=[5, 5],
+        layer_strides=[1, 2],
+        out_channels=[128, 256],
+    ),
+    neck=dict(
+        type='SECONDFPN',
+        in_channels=[128, 256],
+        upsample_strides=[1, 2],
+        out_channels=[256, 256],
+    ),
+    bbox_head=dict(
+        type='Anchor3DHead',
+        num_classes=3,
+        in_channels=512,
+        feat_channels=512,
+        use_direction_classifier=True,
+        anchor_generator=dict(
+            type='Anchor3DRangeGenerator',
+            ranges=[
+                [0, -40.0, -0.6, 70.4, 40.0, -0.6],
+                [0, -40.0, -0.6, 70.4, 40.0, -0.6],
+                [0, -40.0, -1.78, 70.4, 40.0, -1.78],
+            ],
+            sizes=[[0.6, 0.8, 1.73], [0.6, 1.76, 1.73], [1.6, 3.9, 1.56]],
+            rotations=[0, 1.57],
+            reshape_out=False),
+        diff_rad_by_sin=True,
+        bbox_coder=dict(type='DeltaXYZWLHRBBoxCoder'),
+        loss_cls=dict(
+            type='FocalLoss',
+            use_sigmoid=True,
+            gamma=2.0,
+            alpha=0.25,
+            loss_weight=1.0),
+        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=2.0),
+        loss_dir=dict(
+            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2),
+    ),
+)
+# model training and testing settings
+train_cfg = dict(
+    assigner=[
+        dict(  # for Pedestrian
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.35,
+            neg_iou_thr=0.2,
+            min_pos_iou=0.2,
+            ignore_iof_thr=-1),
+        dict(  # for Cyclist
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.35,
+            neg_iou_thr=0.2,
+            min_pos_iou=0.2,
+            ignore_iof_thr=-1),
+        dict(  # for Car
+            type='MaxIoUAssigner',
+            iou_calculator=dict(type='BboxOverlapsNearest3D'),
+            pos_iou_thr=0.6,
+            neg_iou_thr=0.45,
+            min_pos_iou=0.45,
+            ignore_iof_thr=-1),
+    ],
+    allowed_border=0,
+    pos_weight=-1,
+    debug=False)
+test_cfg = dict(
+    use_rotate_nms=True,
+    nms_across_levels=False,
+    nms_thr=0.01,
+    score_thr=0.1,
+    min_bbox_size=0,
+    nms_pre=100,
+    max_num=50)
+
+# dataset settings
+dataset_type = 'KittiDataset'
+data_root = 'data/kitti/'
+class_names = ['Pedestrian', 'Cyclist', 'Car']
+input_modality = dict(use_lidar=True, use_camera=False)
+db_sampler = dict(
+    data_root=data_root,
+    info_path=data_root + 'kitti_dbinfos_train.pkl',
+    rate=1.0,
+    object_rot_range=[0.0, 0.0],
+    prepare=dict(
+        filter_by_difficulty=[-1],
+        filter_by_min_points=dict(
+            Car=5,
+            Pedestrian=10,
+            Cyclist=10,
+        )),
+    classes=class_names,
+    sample_groups=dict(
+        Car=12,
+        Pedestrian=6,
+        Cyclist=6,
+    ))
+file_client_args = dict(backend='disk')
+# file_client_args = dict(
+#     backend='petrel', path_mapping=dict(data='s3://kitti_data/'))
+
+train_pipeline = [
+    dict(
+        type='LoadPointsFromFile',
+        load_dim=4,
+        use_dim=4,
+        file_client_args=file_client_args),
+    dict(
+        type='LoadAnnotations3D',
+        with_bbox_3d=True,
+        with_label_3d=True,
+        file_client_args=file_client_args),
+    dict(type='ObjectSample', db_sampler=db_sampler),
+    dict(
+        type='ObjectNoise',
+        num_try=100,
+        loc_noise_std=[1.0, 1.0, 0.5],
+        global_rot_range=[0.0, 0.0],
+        rot_uniform_noise=[-0.78539816, 0.78539816]),
+    dict(type='RandomFlip3D', flip_ratio=0.5),
+    dict(
+        type='GlobalRotScale',
+        rot_uniform_noise=[-0.78539816, 0.78539816],
+        scaling_uniform_noise=[0.95, 1.05]),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
+    dict(type='PointShuffle'),
+    dict(type='DefaultFormatBundle3D', class_names=class_names),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d']),
+]
+test_pipeline = [
+    dict(
+        type='LoadPointsFromFile',
+        load_dim=4,
+        use_dim=4,
+        file_client_args=file_client_args),
+    dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
+    dict(
+        type='DefaultFormatBundle3D',
+        class_names=class_names,
+        with_label=False),
+    dict(type='Collect3D', keys=['points']),
+]
+
+data = dict(
+    samples_per_gpu=6,
+    workers_per_gpu=4,
+    train=dict(
+        type='RepeatDataset',
+        times=2,
+        dataset=dict(
+            type=dataset_type,
+            data_root=data_root,
+            ann_file=data_root + 'kitti_infos_train.pkl',
+            split='training',
+            pts_prefix='velodyne_reduced',
+            pipeline=train_pipeline,
+            modality=input_modality,
+            classes=class_names,
+            test_mode=False)),
+    val=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True),
+    test=dict(
+        type=dataset_type,
+        data_root=data_root,
+        ann_file=data_root + 'kitti_infos_val.pkl',
+        split='training',
+        pts_prefix='velodyne_reduced',
+        pipeline=test_pipeline,
+        modality=input_modality,
+        classes=class_names,
+        test_mode=True))
+# optimizer
+lr = 0.0018  # max learning rate
+optimizer = dict(type='AdamW', lr=lr, betas=(0.95, 0.99), weight_decay=0.01)
+optimizer_config = dict(grad_clip=dict(max_norm=10, norm_type=2))
+lr_config = dict(
+    policy='cyclic',
+    target_ratio=(10, 1e-4),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+momentum_config = dict(
+    policy='cyclic',
+    target_ratio=(0.85 / 0.95, 1),
+    cyclic_times=1,
+    step_ratio_up=0.4,
+)
+checkpoint_config = dict(interval=1)
+evaluation = dict(interval=1)
+# yapf:disable
+log_config = dict(
+    interval=50,
+    hooks=[
+        dict(type='TextLoggerHook'),
+        dict(type='TensorboardLoggerHook')
+    ])
+# yapf:enable
+# runtime settings
+total_epochs = 40
+dist_params = dict(backend='nccl')
+log_level = 'INFO'
+work_dir = './work_dirs/sec_secfpn_80e'
+load_from = None
+resume_from = None
+workflow = [('train', 1)]

configs/kitti/hv_second_secfpn_6x8_80e_kitti-3d-car.py

@@ -70,7 +70,7 @@ test_cfg = dict(
     use_rotate_nms=True,
     nms_across_levels=False,
     nms_thr=0.01,
-    score_thr=0.3,
+    score_thr=0.1,
     min_bbox_size=0,
     nms_pre=100,
     max_num=50)
@@ -79,15 +79,7 @@ test_cfg = dict(
 dataset_type = 'KittiDataset'
 data_root = 'data/kitti/'
 class_names = ['Car']
-img_norm_cfg = dict(
-    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
-input_modality = dict(
-    use_lidar=False,
-    use_lidar_reduced=True,
-    use_depth=False,
-    use_lidar_intensity=True,
-    use_camera=False,
-)
+input_modality = dict(use_lidar=True, use_camera=False)
 db_sampler = dict(
     data_root=data_root,
     info_path=data_root + 'kitti_dbinfos_train.pkl',

configs/nus/hv_pointpillars_fpn_sbn-all_4x8_2x_nus-3d.py

@@ -108,25 +108,12 @@ test_cfg = dict(
 # dataset settings
 dataset_type = 'NuScenesDataset'
 data_root = 'data/nuscenes/'
-img_norm_cfg = dict(
-    mean=[103.530, 116.280, 123.675], std=[1.0, 1.0, 1.0], to_rgb=False)
 input_modality = dict(
     use_lidar=True,
-    use_depth=False,
-    use_lidar_intensity=True,
     use_camera=False,
-)
-db_sampler = dict(
-    data_root=data_root,
-    info_path=data_root + 'nuscenes_dbinfos_train.pkl',
-    rate=1.0,
-    object_rot_range=[0.0, 0.0],
-    prepare=dict(),
-    sample_groups=dict(
-        bus=4,
-        trailer=4,
-        truck=4,
-    ))
+    use_radar=False,
+    use_map=False,
+    use_external=False)
 file_client_args = dict(backend='disk')
 # file_client_args = dict(
 #     backend='petrel',
@@ -155,7 +142,7 @@ train_pipeline = [
     dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='PointShuffle'),
     dict(type='DefaultFormatBundle3D', class_names=class_names),
-    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d']),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
     dict(
@@ -173,7 +160,7 @@ test_pipeline = [
         type='DefaultFormatBundle3D',
         class_names=class_names,
         with_label=False),
-    dict(type='Collect3D', keys=['points']),
+    dict(type='Collect3D', keys=['points'])
 ]
 
 data = dict(

configs/nus/hv_pointpillars_secfpn_sbn-all_4x8_2x_nus-3d.py

@@ -8,11 +8,10 @@ class_names = [
 model = dict(
     type='MVXFasterRCNNV2',
     pts_voxel_layer=dict(
-        max_num_points=64,  # max_points_per_voxel
-        point_cloud_range=point_cloud_range,  # velodyne coordinates, x, y, z
+        max_num_points=64,
+        point_cloud_range=point_cloud_range,
         voxel_size=voxel_size,
-        max_voxels=(30000, 40000),  # (training, testing) max_coxels
-    ),
+        max_voxels=(30000, 40000)),
     pts_voxel_encoder=dict(
         type='HardVFE',
         in_channels=4,
@@ -24,25 +23,20 @@ model = dict(
         point_cloud_range=point_cloud_range,
         norm_cfg=dict(type='naiveSyncBN1d', eps=1e-3, momentum=0.01)),
     pts_middle_encoder=dict(
-        type='PointPillarsScatter',
-        in_channels=64,
-        output_shape=[400, 400],  # checked from PointCloud3D
-    ),
+        type='PointPillarsScatter', in_channels=64, output_shape=[400, 400]),
     pts_backbone=dict(
         type='SECOND',
         in_channels=64,
         norm_cfg=dict(type='naiveSyncBN2d', eps=1e-3, momentum=0.01),
         layer_nums=[3, 5, 5],
         layer_strides=[2, 2, 2],
-        out_channels=[64, 128, 256],
-    ),
+        out_channels=[64, 128, 256]),
     pts_neck=dict(
         type='SECONDFPN',
         norm_cfg=dict(type='naiveSyncBN2d', eps=1e-3, momentum=0.01),
         in_channels=[64, 128, 256],
         upsample_strides=[1, 2, 4],
-        out_channels=[128, 128, 128],
-    ),
+        out_channels=[128, 128, 128]),
     pts_bbox_head=dict(
         type='Anchor3DHead',
         num_classes=10,
@@ -108,34 +102,17 @@ test_cfg = dict(
         nms_thr=0.2,
         score_thr=0.05,
         min_bbox_size=0,
-        max_num=500
-        # soft-nms is also supported for rcnn testing
-        # e.g., nms=dict(type='soft_nms', iou_thr=0.5, min_score=0.05)
-    ))
+        max_num=500))
 
 # dataset settings
 dataset_type = 'NuScenesDataset'
 data_root = 'data/nuscenes/'
-img_norm_cfg = dict(
-    mean=[103.530, 116.280, 123.675], std=[1.0, 1.0, 1.0], to_rgb=False)
 input_modality = dict(
     use_lidar=True,
-    use_depth=False,
-    use_lidar_intensity=True,
     use_camera=False,
-)
-db_sampler = dict(
-    data_root=data_root,
-    info_path=data_root + 'nuscenes_dbinfos_train.pkl',
-    rate=1.0,
-    object_rot_range=[0.0, 0.0],
-    prepare=dict(),
-    classes=class_names,
-    sample_groups=dict(
-        bus=4,
-        trailer=4,
-        truck=4,
-    ))
+    use_radar=False,
+    use_map=False,
+    use_external=False)
 file_client_args = dict(backend='disk')
 # file_client_args = dict(
 #     backend='petrel',
@@ -164,7 +141,7 @@ train_pipeline = [
     dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='PointShuffle'),
     dict(type='DefaultFormatBundle3D', class_names=class_names),
-    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d']),
+    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
     dict(
@@ -182,7 +159,7 @@ test_pipeline = [
         type='DefaultFormatBundle3D',
         class_names=class_names,
         with_label=False),
-    dict(type='Collect3D', keys=['points']),
+    dict(type='Collect3D', keys=['points'])
 ]
 
 data = dict(

mmdet3d/datasets/__init__.py

-from mmdet.datasets.builder import DATASETS
-from .builder import build_dataset
+from mmdet.datasets.builder import DATASETS, build_dataloader, build_dataset
 from .custom_3d import Custom3DDataset
-from .dataset_wrappers import RepeatFactorDataset
 from .kitti2d_dataset import Kitti2DDataset
 from .kitti_dataset import KittiDataset
-from .loader import DistributedGroupSampler, GroupSampler, build_dataloader
 from .nuscenes_dataset import NuScenesDataset
 from .pipelines import (GlobalRotScale, IndoorFlipData, IndoorGlobalRotScale,
                         IndoorPointSample, IndoorPointsColorJitter,
@@ -18,6 +15,7 @@ from .sunrgbd_dataset import SUNRGBDDataset
 __all__ = [
     'KittiDataset', 'GroupSampler', 'DistributedGroupSampler',
     'build_dataloader', 'RepeatFactorDataset', 'DATASETS', 'build_dataset',
+    'build_dataloader'
     'CocoDataset', 'Kitti2DDataset', 'NuScenesDataset', 'ObjectSample',
     'RandomFlip3D', 'ObjectNoise', 'GlobalRotScale', 'PointShuffle',
     'ObjectRangeFilter', 'PointsRangeFilter', 'Collect3D',

mmdet3d/datasets/builder.py

-import copy
-
-from mmcv.utils import build_from_cfg
-
-from mmdet.datasets import DATASETS, ConcatDataset, RepeatDataset
-from .dataset_wrappers import RepeatFactorDataset
-
-
-def _concat_dataset(cfg, default_args=None):
-    ann_files = cfg['ann_file']
-    img_prefixes = cfg.get('img_prefix', None)
-    seg_prefixes = cfg.get('seg_prefix', None)
-    proposal_files = cfg.get('proposal_file', None)
-
-    datasets = []
-    num_dset = len(ann_files)
-    for i in range(num_dset):
-        data_cfg = copy.deepcopy(cfg)
-        data_cfg['ann_file'] = ann_files[i]
-        if isinstance(img_prefixes, (list, tuple)):
-            data_cfg['img_prefix'] = img_prefixes[i]
-        if isinstance(seg_prefixes, (list, tuple)):
-            data_cfg['seg_prefix'] = seg_prefixes[i]
-        if isinstance(proposal_files, (list, tuple)):
-            data_cfg['proposal_file'] = proposal_files[i]
-        datasets.append(build_dataset(data_cfg, default_args))
-
-    return ConcatDataset(datasets)
-
-
-def build_dataset(cfg, default_args=None):
-    if isinstance(cfg, (list, tuple)):
-        dataset = ConcatDataset([build_dataset(c, default_args) for c in cfg])
-    elif cfg['type'] == 'RepeatDataset':
-        dataset = RepeatDataset(
-            build_dataset(cfg['dataset'], default_args), cfg['times'])
-    elif cfg['type'] == 'RepeatFactorDataset':
-        dataset = RepeatFactorDataset(
-            build_dataset(cfg['dataset'], default_args), cfg['repeat_thr'])
-    elif isinstance(cfg.get('ann_file'), (list, tuple)):
-        dataset = _concat_dataset(cfg, default_args)
-    else:
-        dataset = build_from_cfg(cfg, DATASETS, default_args)
-
-    return dataset

mmdet3d/datasets/dataset_wrappers.py

-import math
-from collections import defaultdict
-
-import numpy as np
-
-from mmdet.datasets import DATASETS
-
-
-# Modified from https://github.com/facebookresearch/detectron2/blob/41d475b75a230221e21d9cac5d69655e3415e3a4/detectron2/data/samplers/distributed_sampler.py#L57 # noqa
-@DATASETS.register_module()
-class RepeatFactorDataset(object):
-    """A wrapper of repeated dataset with repeat factor.
-
-    Suitable for training on class imbalanced datasets like LVIS. In each
-    epoch, an image may appear multiple times based on its "repeat factor".
-    The repeat factor for an image is a function of the frequency the rarest
-    category labeled in that image. The "frequency of category c" in [0, 1]
-    is defined as the fraction of images in the training set (without repeats)
-    in which category c appears.
-    This wrapper will finally be merged into LVIS dataset.
-
-    See https://arxiv.org/abs/1908.03195 (>= v2) Appendix B.2.
-    Args:
-        dataset (:obj:`Dataset`): The dataset to be repeated.
-        repeat_thr (float): frequency threshold below which data is repeated.
-    """
-
-    def __init__(self, dataset, repeat_thr):
-        self.dataset = dataset
-        self.repeat_thr = repeat_thr
-        self.CLASSES = dataset.CLASSES
-
-        repeat_factors = self._get_repeat_factors(dataset, repeat_thr)
-        repeat_indices = []
-        for dataset_index, repeat_factor in enumerate(repeat_factors):
-            repeat_indices.extend([dataset_index] * math.ceil(repeat_factor))
-        self.repeat_indices = repeat_indices
-
-        flags = []
-        if hasattr(self.dataset, 'flag'):
-            for flag, repeat_factor in zip(self.dataset.flag, repeat_factors):
-                flags.extend([flag] * int(math.ceil(repeat_factor)))
-            assert len(flags) == len(repeat_indices)
-        self.flag = np.asarray(flags, dtype=np.uint8)
-
-    def _get_repeat_factors(self, dataset, repeat_thr):
-        # 1. For each category c, compute the fraction # of images
-        # that contain it: f(c)
-        category_freq = defaultdict(int)
-        for idx, img_info in enumerate(dataset.data_infos):
-            if 'category_ids' in img_info:
-                cat_ids = set(img_info['category_ids'])
-            elif 'gt_names' in img_info:
-                cat_ids = set([
-                    gt for gt in img_info['gt_names']
-                    if gt in dataset.class_names
-                ])
-            else:
-                labels = dataset.get_ann_info(idx)['labels']
-                cat_ids = set([label for label in labels])
-            for cat_id in cat_ids:
-                category_freq[cat_id] += 1
-        num_images = len(dataset)
-        for k, v in category_freq.items():
-            category_freq[k] = v / num_images
-
-        # 2. For each category c, compute the category-level repeat factor:
-        #    r(c) = max(1, sqrt(t / f(c)))
-        category_repeat = {
-            cat_id: max(1.0, math.sqrt(repeat_thr / cat_freq))
-            for cat_id, cat_freq in category_freq.items()
-        }
-
-        # 3. For each image I, compute the image-level repeat factor:
-        #    r(I) = max_{c in I} r(c)
-        repeat_factors = []
-        for idx, img_info in enumerate(dataset.data_infos):
-            if 'category_ids' in img_info:
-                cat_ids = set(img_info['category_ids'])
-            elif 'gt_names' in img_info:
-                cat_ids = set([
-                    gt for gt in img_info['gt_names']
-                    if gt in dataset.class_names
-                ])
-            else:
-                labels = dataset.get_ann_info(idx)['labels']
-                cat_ids = set([label for label in labels])
-
-            if len(cat_ids) == 0:
-                repeat_factor = 1
-            else:
-                repeat_factor = max(
-                    {category_repeat[cat_id]
-                     for cat_id in cat_ids})
-            repeat_factors.append(repeat_factor)
-        return repeat_factors
-
-    def __getitem__(self, idx):
-        ori_index = self.repeat_indices[idx]
-        return self.dataset[ori_index]
-
-    def __len__(self):
-        return len(self.repeat_indices)

mmdet3d/datasets/kitti2d_dataset.py

@@ -53,7 +53,7 @@ class Kitti2DDataset(CustomDataset):
         self.data_infos = mmcv.load(ann_file)
         self.cat2label = {
             cat_name: i
-            for i, cat_name in enumerate(self.class_names)
+            for i, cat_name in enumerate(self.CLASSES)
         }
         return self.data_infos
 
@@ -107,14 +107,6 @@ class Kitti2DDataset(CustomDataset):
         self.pre_pipeline(results)
         return self.pipeline(results)
 
-    def _set_group_flag(self):
-        """Set flag according to image aspect ratio.
-        Images with aspect ratio greater than 1 will be set as group 1,
-        otherwise group 0.
-        In kitti's pcd, they are all the same, thus are all zeros
-        """
-        self.flag = np.zeros(len(self), dtype=np.uint8)
-
     def drop_arrays_by_name(self, gt_names, used_classes):
         inds = [i for i, x in enumerate(gt_names) if x not in used_classes]
         inds = np.array(inds, dtype=np.int64)