add configs for PartA2 and PV-RCNN, and add setup.py for compiling

pcdet/config.py

+from easydict import EasyDict
+from pathlib import Path
+import yaml
+
+
+def log_config_to_file(cfg, pre='cfg', logger=None):
+    for key, val in cfg.items():
+        if isinstance(cfg[key], EasyDict):
+            logger.info('\n%s.%s = edict()' % (pre, key))
+            log_config_to_file(cfg[key], pre=pre + '.' + key, logger=logger)
+            continue
+        logger.info('%s.%s: %s' % (pre, key, val))
+
+
+def cfg_from_list(cfg_list, config):
+    """Set config keys via list (e.g., from command line)."""
+    from ast import literal_eval
+    assert len(cfg_list) % 2 == 0
+    for k, v in zip(cfg_list[0::2], cfg_list[1::2]):
+        key_list = k.split('.')
+        d = config
+        for subkey in key_list[:-1]:
+            assert subkey in d, 'NotFoundKey: %s' % subkey
+            d = d[subkey]
+        subkey = key_list[-1]
+        assert subkey in d, 'NotFoundKey: %s' % subkey
+        try:
+            value = literal_eval(v)
+        except:
+            value = v
+
+        if type(value) != type(d[subkey]) and isinstance(d[subkey], EasyDict):
+            key_val_list = value.split(',')
+            for src in key_val_list:
+                cur_key, cur_val = src.split(':')
+                val_type = type(d[subkey][cur_key])
+                cur_val = val_type(cur_val)
+                d[subkey][cur_key] = cur_val
+        elif type(value) != type(d[subkey]) and isinstance(d[subkey], list):
+            val_list = value.split(',')
+            for k, x in enumerate(val_list):
+                val_list[k] = type(d[subkey][0])(x)
+            d[subkey] = val_list
+        else:
+            assert type(value) == type(d[subkey]), \
+                'type {} does not match original type {}'.format(type(value), type(d[subkey]))
+            d[subkey] = value
+
+
+def merge_new_config(config, new_config):
+    if '_BASE_CONFIG_' in new_config:
+        with open(new_config['_BASE_CONFIG_'], 'r') as f:
+            try:
+                yaml_config = yaml.load(f)
+            except:
+                yaml_config = yaml.load(f, Loader=yaml.FullLoader)
+        config.update(EasyDict(yaml_config))
+
+    for key, val in new_config.items():
+        if not isinstance(val, dict):
+            config[key] = val
+            continue
+        if key not in config:
+            config[key] = EasyDict()
+        merge_new_config(config[key], val)
+
+    return config
+
+
+def cfg_from_yaml_file(cfg_file, config):
+    with open(cfg_file, 'r') as f:
+        try:
+            new_config = yaml.load(f)
+        except:
+            new_config = yaml.load(f, Loader=yaml.FullLoader)
+
+        merge_new_config(config=config, new_config=new_config)
+
+    return config
+
+
+cfg = EasyDict()
+cfg.ROOT_DIR = (Path(__file__).resolve().parent / '../').resolve()
+cfg.LOCAL_RANK = 0

setup.py

+import os
+from setuptools import setup, find_packages
+import subprocess
+from torch.utils.cpp_extension import BuildExtension, CUDAExtension
+
+
+def get_git_commit_number():
+    if not os.path.exists('.git'):
+        return '0000000'
+
+    cmd_out = subprocess.run(['git', 'rev-parse', 'HEAD'], stdout=subprocess.PIPE)
+    git_commit_number = cmd_out.stdout.decode('utf-8')[:7]
+    return git_commit_number
+
+
+def make_cuda_ext(name, module, sources):
+    cuda_ext = CUDAExtension(
+        name='%s.%s' % (module, name),
+        sources=[os.path.join(*module.split('.'), src) for src in sources]
+    )
+    return cuda_ext
+
+
+def write_version_to_file(version, target_file):
+    with open(target_file, 'w') as f:
+        print('__version__ = "%s"' % version, file=f)
+
+
+if __name__ == '__main__':
+    version = '0.2.0+%s' % get_git_commit_number()
+    write_version_to_file(version, 'pcdet/version.py')
+
+    setup(
+        name='pcdet',
+        version=version,
+        description='PCDet is a general codebase for 3D object detection from point cloud',
+        install_requires=[
+            'numpy',
+            'torch>=1.1',
+            'spconv==1.0',
+            'numba',
+            'tensorboardX',
+            'easydict',
+            'pyyaml'
+        ],
+        author='Shaoshuai Shi',
+        author_email='shaoshuaics@gmail.com',
+        license='Apache License 2.0',
+        packages=find_packages(exclude=['tools', 'data', 'output']),
+        cmdclass={'build_ext': BuildExtension},
+        ext_modules=[
+            make_cuda_ext(
+                name='iou3d_nms_cuda',
+                module='pcdet.ops.iou3d_nms',
+                sources=[
+                    'src/iou3d_cpu.cpp',
+                    'src/iou3d_nms_api.cpp',
+                    'src/iou3d_nms.cpp',
+                    'src/iou3d_nms_kernel.cu',
+                ]
+            ),
+            make_cuda_ext(
+                name='roiaware_pool3d_cuda',
+                module='pcdet.ops.roiaware_pool3d',
+                sources=[
+                    'src/roiaware_pool3d.cpp',
+                    'src/roiaware_pool3d_kernel.cu',
+                ]
+            ),
+            make_cuda_ext(
+                name='pointnet2_stack_cuda',
+                module='pcdet.ops.pointnet2.pointnet2_stack',
+                sources=[
+                    'src/pointnet2_api.cpp',
+                    'src/ball_query.cpp',
+                    'src/ball_query_gpu.cu',
+                    'src/group_points.cpp',
+                    'src/group_points_gpu.cu',
+                    'src/sampling.cpp',
+                    'src/sampling_gpu.cu',
+                ],
+            ),
+        ],
+    )
+

tools/cfgs/dataset_configs/kitti_dataset.yaml

+DATASET: 'KittiDataset'
+DATA_PATH: '../data/kitti'
+
+POINT_CLOUD_RANGE: [0, -40, -3, 70.4, 40, 1]
+
+DATA_SPLIT: {
+    'train': train,
+    'test': val
+}
+
+INFO_PATH: {
+    'train': [kitti_infos_train.pkl],
+    'test': [kitti_infos_val.pkl],
+}
+
+FOV_POINTS_ONLY: True
+
+
+DATA_AUGMENTOR:
+    - NAME: gt_sampling
+      USE_ROAD_PLANE: True
+      DB_INFO_PATH:
+          - kitti_dbinfos_train.pkl
+      PREPARE: {
+         filter_by_min_points: ['Car:5', 'Pedestrian:5', 'Cyclist:5'],
+         filter_by_difficulty: [-1],
+      }
+
+      SAMPLE_GROUPS: ['Car:20','Pedestrian:15', 'Cyclist:15']
+      NUM_POINT_FEATURES: 4
+      DATABASE_WITH_FAKELIDAR: True
+      REMOVE_EXTRA_WIDTH: [0.0, 0.0, 0.0]
+      LIMIT_WHOLE_SCENE: True
+
+    - NAME: random_world_flip
+      ALONG_AXIS_LIST: ['x']
+
+    - NAME: random_world_rotation
+      WORLD_ROT_ANGLE: [-0.78539816, 0.78539816]
+
+    - NAME: random_world_scaling
+      WORLD_SCALE_RANGE: [0.95, 1.05]
+
+
+POINT_FEATURE_ENCODING: {
+    encoding_type: absolute_coordinates_encoding,
+    used_feature_list: ['x', 'y', 'z', 'intensity'],
+    src_feature_list: ['x', 'y', 'z', 'intensity'],
+}
+
+
+DATA_PROCESSOR:
+    - NAME: mask_points_and_boxes_outside_range
+      REMOVE_OUTSIDE_BOXES: True
+
+    - NAME: shuffle_points
+      SHUFFLE_ENABLED: {
+        'train': True,
+        'test': False
+      }
+
+    - NAME: transform_points_to_voxels
+      VOXEL_SIZE: [0.05, 0.05, 0.1]
+      MAX_POINTS_PER_VOXEL: 5
+      MAX_NUMBER_OF_VOXELS: {
+        'train': 16000,
+        'test': 40000
+      }

tools/cfgs/kitti_models/PartA2.yaml

+CLASS_NAMES: ['Car', 'Pedestrian', 'Cyclist']
+
+DATA_CONFIG:
+    _BASE_CONFIG_: cfgs/dataset_configs/kitti_dataset.yaml
+
+
+MODEL:
+    NAME: PartA2Net
+
+    VFE:
+        NAME: MeanVFE
+
+    BACKBONE_3D:
+        NAME: UNetV2
+
+    MAP_TO_BEV:
+        NAME: HeightCompression
+        NUM_BEV_FEATURES: 256
+
+    BACKBONE_2D:
+        NAME: BaseBEVBackbone
+
+        LAYER_NUMS: [5, 5]
+        LAYER_STRIDES: [1, 2]
+        NUM_FILTERS: [128, 256]
+        UPSAMPLE_STRIDES: [1, 2]
+        NUM_UPSAMPLE_FILTERS: [256, 256]
+
+    DENSE_HEAD:
+        NAME: AnchorHeadSingle
+        CLASS_AGNOSTIC: False
+
+        USE_DIRECTION_CLASSIFIER: True
+        DIR_OFFSET: 0.78539
+        DIR_LIMIT_OFFSET: 0.0
+        NUM_DIR_BINS: 2
+
+        ANCHOR_GENERATOR_CONFIG: [
+            {
+                'anchor_sizes': [[3.9, 1.6, 1.56]],
+                'anchor_rotations': [0, 1.57],
+                'anchor_bottom_heights': [-1.78],
+                'align_center': False,
+                'feature_map_stride': 8
+            },
+            {
+                'anchor_sizes': [[0.8, 0.6, 1.73]],
+                'anchor_rotations': [0, 1.57],
+                'anchor_bottom_heights': [-1.78],
+                'align_center': False,
+                'feature_map_stride': 8
+            },
+            {
+                'anchor_sizes': [[1.76, 0.6, 1.73]],
+                'anchor_rotations': [0, 1.57],
+                'anchor_bottom_heights': [-1.78],
+                'align_center': False,
+                'feature_map_stride': 8
+            }
+        ]
+
+        TARGET_ASSIGNER_CONFIG:
+            NAME: AxisAlignedTargetAssigner
+            POS_FRACTION: -1.0
+            SAMPLE_SIZE: 512
+            MATCHED_THRESHOLDS: [0.6, 0.5, 0.5]
+            UNMATCHED_THRESHOLDS: [0.45, 0.35, 0.35]
+            NORM_BY_NUM_EXAMPLES: False
+            MATCH_HEIGHT: False
+            BOX_CODER: ResidualCoder
+
+        LOSS_CONFIG:
+            LOSS_WEIGHTS: {
+                'cls_weight': 1.0,
+                'loc_weight': 2.0,
+                'dir_weight': 0.2,
+                'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
+            }
+
+    POINT_HEAD:
+        NAME: PointIntraPartOffsetHead
+        CLS_FC: []
+        PART_FC: []
+        CLASS_AGNOSTIC: True
+        TARGET_CONFIG:
+            GT_EXTRA_WIDTH: [0.2, 0.2, 0.2]
+        LOSS_CONFIG:
+            LOSS_REG: smooth-l1
+            LOSS_WEIGHTS: {
+                'point_cls_weight': 1.0,
+                'point_part_weight': 1.0
+            }
+
+    ROI_HEAD:
+        NAME: PartA2FCHead
+        CLASS_AGNOSTIC: True
+
+        SHARED_FC: [256, 256, 256]
+        CLS_FC: [256, 256]
+        REG_FC: [256, 256]
+        DP_RATIO: 0.3
+
+        SEG_MASK_SCORE_THRESH: 0.3
+
+        NMS_CONFIG:
+            TRAIN:
+                NMS_TYPE: nms_gpu
+                MULTI_CLASSES_NMS: False
+                NMS_PRE_MAXSIZE: 9000
+                NMS_POST_MAXSIZE: 512
+                NMS_THRESH: 0.8
+            TEST:
+                NMS_TYPE: nms_gpu
+                MULTI_CLASSES_NMS: False
+                NMS_PRE_MAXSIZE: 1024
+                NMS_POST_MAXSIZE: 100
+                NMS_THRESH: 0.7
+
+        ROI_AWARE_POOL:
+            POOL_SIZE: 12
+            NUM_FEATURES: 128
+            MAX_POINTS_PER_VOXEL: 128
+
+        TARGET_CONFIG:
+            BOX_CODER: ResidualCoder
+            ROI_PER_IMAGE: 128
+            FG_RATIO: 0.5
+
+            SAMPLE_ROI_BY_EACH_CLASS: True
+            CLS_SCORE_TYPE: roi_iou
+
+            CLS_FG_THRESH: 0.75
+            CLS_BG_THRESH: 0.25
+            CLS_BG_THRESH_LO: 0.1
+            HARD_BG_RATIO: 0.8
+
+            REG_FG_THRESH: 0.65
+
+        LOSS_CONFIG:
+            CLS_LOSS: BinaryCrossEntropy
+            REG_LOSS: smooth-l1
+            CORNER_LOSS_REGULARIZATION: True
+            LOSS_WEIGHTS: {
+                'rcnn_cls_weight': 1.0,
+                'rcnn_reg_weight': 1.0,
+                'rcnn_corner_weight': 1.0,
+                'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
+            }
+
+    POST_PROCESSING:
+        RECALL_THRESH_LIST: [0.3, 0.5, 0.7]
+        SCORE_THRESH: 0.1
+        OUTPUT_RAW_SCORE: False
+
+        EVAL_METRIC: kitti
+
+        NMS_CONFIG:
+            MULTI_CLASSES_NMS: False
+            NMS_TYPE: nms_gpu
+            NMS_THRESH: 0.1
+            NMS_PRE_MAXSIZE: 4096
+            NMS_POST_MAXSIZE: 500
+
+
+OPTIMIZATION:
+    OPTIMIZER: adam_onecycle
+    LR: 0.01
+    WEIGHT_DECAY: 0.01
+    MOMENTUM: 0.9
+
+    MOMS: [0.95, 0.85]
+    PCT_START: 0.4
+    DIV_FACTOR: 10
+    DECAY_STEP_LIST: [35, 45]
+    LR_DECAY: 0.1
+    LR_CLIP: 0.0000001
+
+    LR_WARMUP: False
+    WARMUP_EPOCH: 1
+
+    GRAD_NORM_CLIP: 10

tools/cfgs/kitti_models/pv_rcnn.yaml

+CLASS_NAMES: ['Car', 'Pedestrian', 'Cyclist']
+
+DATA_CONFIG:
+    _BASE_CONFIG_: cfgs/dataset_configs/kitti_dataset.yaml
+
+
+MODEL:
+    NAME: PVRCNN
+
+    VFE:
+        NAME: MeanVFE
+
+    BACKBONE_3D:
+        NAME: VoxelBackBone8x
+
+    MAP_TO_BEV:
+        NAME: HeightCompression
+        NUM_BEV_FEATURES: 256
+
+    BACKBONE_2D:
+        NAME: BaseBEVBackbone
+
+        LAYER_NUMS: [5, 5]
+        LAYER_STRIDES: [1, 2]
+        NUM_FILTERS: [128, 256]
+        UPSAMPLE_STRIDES: [1, 2]
+        NUM_UPSAMPLE_FILTERS: [256, 256]
+
+    DENSE_HEAD:
+        NAME: AnchorHeadSingle
+        CLASS_AGNOSTIC: False
+
+        USE_DIRECTION_CLASSIFIER: True
+        DIR_OFFSET: 0.78539
+        DIR_LIMIT_OFFSET: 0.0
+        NUM_DIR_BINS: 2
+
+        ANCHOR_GENERATOR_CONFIG: [
+            {
+                'anchor_sizes': [[3.9, 1.6, 1.56]],
+                'anchor_rotations': [0, 1.57],
+                'anchor_bottom_heights': [-1.78],
+                'align_center': False,
+                'feature_map_stride': 8
+            },
+            {
+                'anchor_sizes': [[0.8, 0.6, 1.73]],
+                'anchor_rotations': [0, 1.57],
+                'anchor_bottom_heights': [-0.6],
+                'align_center': False,
+                'feature_map_stride': 8
+            },
+            {
+                'anchor_sizes': [[1.76, 0.6, 1.73]],
+                'anchor_rotations': [0, 1.57],
+                'anchor_bottom_heights': [-0.6],
+                'align_center': False,
+                'feature_map_stride': 8
+            }
+        ]
+
+        TARGET_ASSIGNER_CONFIG:
+            NAME: AxisAlignedTargetAssigner
+            POS_FRACTION: -1.0
+            SAMPLE_SIZE: 512
+            MATCHED_THRESHOLDS: [0.6, 0.5, 0.5]
+            UNMATCHED_THRESHOLDS: [0.45, 0.35, 0.35]
+            NORM_BY_NUM_EXAMPLES: False
+            MATCH_HEIGHT: False
+            BOX_CODER: ResidualCoder
+
+        LOSS_CONFIG:
+            LOSS_WEIGHTS: {
+                'cls_weight': 1.0,
+                'loc_weight': 2.0,
+                'dir_weight': 0.2,
+                'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
+            }
+
+    PFE:
+        NAME: VoxelSetAbstraction
+        POINT_SOURCE: voxel_centers
+        NUM_KEYPOINTS: 2048
+        NUM_OUTPUT_FEATURES: 128
+        SAMPLE_METHOD: FPS
+
+        FEATURES_SOURCE: ['bev', 'x_conv1', 'x_conv2', 'x_conv3', 'x_conv4', 'raw_points']
+        SA_LAYER:
+            raw_points:
+                MLPS: [[16, 16], [16, 16]]
+                POOL_RADIUS: [0.4, 0.8]
+                NSAMPLE: [16, 16]
+            x_conv1:
+                DOWNSAMPLE_FACTOR: 1
+                MLPS: [[16, 16], [16, 16]]
+                POOL_RADIUS: [0.4, 0.8]
+                NSAMPLE: [16, 16]
+            x_conv2:
+                DOWNSAMPLE_FACTOR: 2
+                MLPS: [[32, 32], [32, 32]]
+                POOL_RADIUS: [0.8, 1.2]
+                NSAMPLE: [16, 32]
+            x_conv3:
+                DOWNSAMPLE_FACTOR: 4
+                MLPS: [[64, 64], [64, 64]]
+                POOL_RADIUS: [1.2, 2.4]
+                NSAMPLE: [16, 32]
+            x_conv4:
+                DOWNSAMPLE_FACTOR: 8
+                MLPS: [[64, 64], [64, 64]]
+                POOL_RADIUS: [2.4, 4.8]
+                NSAMPLE: [16, 32]
+
+    POINT_HEAD:
+        NAME: PointHeadSimple
+        CLS_FC: [256, 256]
+        CLASS_AGNOSTIC: True
+        USE_POINT_FEATURES_BEFORE_FUSION: True
+        TARGET_CONFIG:
+            GT_EXTRA_WIDTH: [0.2, 0.2, 0.2]
+        LOSS_CONFIG:
+            LOSS_REG: smooth-l1
+            LOSS_WEIGHTS: {
+                'point_cls_weight': 1.0,
+            }
+
+    ROI_HEAD:
+        NAME: PVRCNNHead
+        CLASS_AGNOSTIC: True
+
+        SHARED_FC: [256, 256]
+        CLS_FC: [256, 256]
+        REG_FC: [256, 256]
+        DP_RATIO: 0.2
+
+        NMS_CONFIG:
+            TRAIN:
+                NMS_TYPE: nms_gpu
+                MULTI_CLASSES_NMS: False
+                NMS_PRE_MAXSIZE: 9000
+                NMS_POST_MAXSIZE: 512
+                NMS_THRESH: 0.8
+            TEST:
+                NMS_TYPE: nms_gpu
+                MULTI_CLASSES_NMS: False
+                NMS_PRE_MAXSIZE: 1024
+                NMS_POST_MAXSIZE: 100
+                NMS_THRESH: 0.7
+
+        ROI_GRID_POOL:
+            GRID_SIZE: 6
+            MLPS: [[64, 64], [64, 64]]
+            POOL_RADIUS: [0.8, 1.6]
+            NSAMPLE: [16, 16]
+            POOL_METHOD: max_pool
+
+        TARGET_CONFIG:
+            BOX_CODER: ResidualCoder
+            ROI_PER_IMAGE: 128
+            FG_RATIO: 0.5
+
+            SAMPLE_ROI_BY_EACH_CLASS: True
+            CLS_SCORE_TYPE: roi_iou
+
+            CLS_FG_THRESH: 0.75
+            CLS_BG_THRESH: 0.25
+            CLS_BG_THRESH_LO: 0.1
+            HARD_BG_RATIO: 0.8
+
+            REG_FG_THRESH: 0.55
+
+        LOSS_CONFIG:
+            CLS_LOSS: BinaryCrossEntropy
+            REG_LOSS: smooth-l1
+            CORNER_LOSS_REGULARIZATION: True
+            LOSS_WEIGHTS: {
+                'rcnn_cls_weight': 1.0,
+                'rcnn_reg_weight': 1.0,
+                'rcnn_corner_weight': 1.0,
+                'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
+            }
+
+    POST_PROCESSING:
+        RECALL_THRESH_LIST: [0.3, 0.5, 0.7]
+        SCORE_THRESH: 0.1
+        OUTPUT_RAW_SCORE: False
+
+        EVAL_METRIC: kitti
+
+        NMS_CONFIG:
+            MULTI_CLASSES_NMS: False
+            NMS_TYPE: nms_gpu
+            NMS_THRESH: 0.1
+            NMS_PRE_MAXSIZE: 4096
+            NMS_POST_MAXSIZE: 500
+
+
+OPTIMIZATION:
+    OPTIMIZER: adam_onecycle
+    LR: 0.01
+    WEIGHT_DECAY: 0.01
+    MOMENTUM: 0.9
+
+    MOMS: [0.95, 0.85]
+    PCT_START: 0.4
+    DIV_FACTOR: 10
+    DECAY_STEP_LIST: [35, 45]
+    LR_DECAY: 0.1
+    LR_CLIP: 0.0000001
+
+    LR_WARMUP: False
+    WARMUP_EPOCH: 1
+
+    GRAD_NORM_CLIP: 10

tools/train.py

@@ -33,6 +33,7 @@ def parge_config():
     parser.add_argument('--ckpt_save_interval', type=int, default=1, help='number of training epochs')
     parser.add_argument('--local_rank', type=int, default=0, help='local rank for distributed training')
     parser.add_argument('--max_ckpt_save_num', type=int, default=30, help='max number of saved checkpoint')
+    parser.add_argument('--merge_all_iters_to_one_epoch', action='store_true', default=False, help='')
     parser.add_argument('--set', dest='set_cfgs', default=None, nargs=argparse.REMAINDER,
                         help='set extra config keys if needed')