safdnet_20e_nuscenes.yaml

CLASS_NAMES: ['car','truck', 'construction_vehicle', 'bus', 'trailer',
              'barrier', 'motorcycle', 'bicycle', 'pedestrian', 'traffic_cone']

DATA_CONFIG:
    _BASE_CONFIG_: cfgs/dataset_configs/nuscenes_dataset.yaml
    POINT_CLOUD_RANGE: [-54.0, -54.0, -5.0, 54.0, 54.0, 3.0]

    DATA_AUGMENTOR:
        DISABLE_AUG_LIST: ['placeholder']
        AUG_CONFIG_LIST:
          - NAME: gt_sampling
            DB_INFO_PATH:
              - nuscenes_dbinfos_10sweeps_withvelo.pkl

            USE_SHARED_MEMORY: True
            DB_DATA_PATH:
              - nuscenes_10sweeps_withvelo_lidar.npy

            PREPARE: {
              filter_by_min_points: [
                  'car:5','truck:5', 'construction_vehicle:5', 'bus:5', 'trailer:5',
                  'barrier:5', 'motorcycle:5', 'bicycle:5', 'pedestrian:5', 'traffic_cone:5'
              ],
            }

            SAMPLE_GROUPS: [
              'car:2','truck:3', 'construction_vehicle:7', 'bus:4', 'trailer:6',
              'barrier:2', 'motorcycle:6', 'bicycle:6', 'pedestrian:2', 'traffic_cone:2'
            ]

            NUM_POINT_FEATURES: 5
            DATABASE_WITH_FAKELIDAR: False
            REMOVE_EXTRA_WIDTH: [0.0, 0.0, 0.0]
            LIMIT_WHOLE_SCENE: True

          - NAME: random_world_flip
            ALONG_AXIS_LIST: ['x', 'y']

          - NAME: random_world_rotation
            WORLD_ROT_ANGLE: [-0.78539816, 0.78539816]

          - NAME: random_world_scaling
            WORLD_SCALE_RANGE: [0.9, 1.1]

          - NAME: random_world_translation
            NOISE_TRANSLATE_STD: [0.5, 0.5, 0.5]

    DATA_PROCESSOR:
      - NAME: mask_points_and_boxes_outside_range
        REMOVE_OUTSIDE_BOXES: True

      - NAME: shuffle_points
        SHUFFLE_ENABLED: {
          'train': True,
          'test': True
        }

      - NAME: transform_points_to_voxels_placeholder
        VOXEL_SIZE: [0.3, 0.3, 8.0]

MODEL:
    NAME: TransFusion

    VFE:
        NAME: DynPillarVFE
        WITH_DISTANCE: False
        USE_ABSLOTE_XYZ: True
        USE_NORM: True
        NUM_FILTERS: [128, 128]

    BACKBONE_3D:
        NAME: SparseHEDNet2D

        SED_FEATURE_DIM: 128
        SED_NUM_LAYERS: 6
        SED_NUM_SBB: [2, 1, 1]
        SED_DOWN_STRIDE: [1, 2, 2]
        SED_DOWN_KERNEL_SIZE: [3, 3, 3]

        AFD_FEATURE_DIM: 128
        AFD_NUM_LAYERS: 1
        AFD_NUM_SBB: [4, 4, 4]
        AFD_DOWN_STRIDE: [1, 2, 2]
        AFD_DOWN_KERNEL_SIZE: [3, 3, 3]

        AFD: True
        FEATMAP_STRIDE: 2
        DETACH_FEATURE: True
        FG_THRESHOLD: 0.3
        GREOUP_POOLING_KERNEL_SIZE: [9, 15, 7, 7]   # NDS 70.7~71.1
        GROUP_CLASS_NAMES: [
            ['car', 'truck', 'construction_vehicle'],
            ['bus', 'trailer'],
            ['barrier', 'motorcycle', 'bicycle', 'pedestrian', 'traffic_cone'],
        ]

    DENSE_HEAD:
        NAME: SparseTransFusionHead
        CLASS_AGNOSTIC: False
        USE_BIAS_BEFORE_NORM: False
        USE_TENSOR_MASK: True

        INPUT_FEATURES: 128
        NUM_PROPOSALS: 200  # set it to 300 when performing inference on the test set (inference only)
        HIDDEN_CHANNEL: 128
        NUM_CLASSES: 10
        NUM_HEADS: 8
        NMS_KERNEL_SIZE: 3
        FFN_CHANNEL: 256
        DROPOUT: 0.1
        BN_MOMENTUM: 0.1
        ACTIVATION: relu

        NUM_HM_CONV: 2
        SEPARATE_HEAD_CFG:
            HEAD_ORDER: ['center', 'height', 'dim', 'rot', 'vel']
            HEAD_DICT: {
                'center': {'out_channels': 2, 'num_conv': 2},
                'height': {'out_channels': 1, 'num_conv': 2},
                'dim': {'out_channels': 3, 'num_conv': 2},
                'rot': {'out_channels': 2, 'num_conv': 2},
                'vel': {'out_channels': 2, 'num_conv': 2},
            }

        TARGET_ASSIGNER_CONFIG:
            DATASET: nuScenes
            FEATURE_MAP_STRIDE: 2
            GAUSSIAN_OVERLAP: 0.1
            MIN_RADIUS: 2
            HUNGARIAN_ASSIGNER:
                cls_cost: {'gamma': 2.0, 'alpha': 0.25, 'weight': 0.15}
                reg_cost: {'weight': 0.25}
                iou_cost: {'weight': 0.25}

        LOSS_CONFIG:
            LOSS_WEIGHTS: {
                    'cls_weight': 1.0,
                    'bbox_weight': 0.25,
                    'hm_weight': 1.0,
                    'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.2, 0.2]
                }
            LOSS_CLS:
                use_sigmoid: True
                gamma: 2.0
                alpha: 0.25
            LOSS_IOU: False
            LOSS_IOU_REG: False

        POST_PROCESSING:
            SCORE_THRESH: 0.0
            POST_CENTER_RANGE: [-61.2, -61.2, -10.0, 61.2, 61.2, 10.0]
            USE_IOU_TO_RECTIFY_SCORE: False
            IOU_RECTIFIER: [0.5]

        NMS_CONFIG:
            NMS_TYPE: nms_gpu
            NMS_THRESH: 0.2
            NMS_PRE_MAXSIZE: 1000
            NMS_POST_MAXSIZE: 100
            SCORE_THRES: 0.

    POST_PROCESSING:
        RECALL_THRESH_LIST: [0.3, 0.5, 0.7]
        SCORE_THRESH: 0.1
        OUTPUT_RAW_SCORE: False
        EVAL_METRIC: kitti

OPTIMIZATION:
    BATCH_SIZE_PER_GPU: 2
    NUM_EPOCHS: 20

    OPTIMIZER: adam_onecycle
    LR: 0.003
    WEIGHT_DECAY: 0.05
    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: 35
    LOSS_SCALE_FP16: 4.0

HOOK:
    DisableAugmentationHook:
        DISABLE_AUG_LIST: ['gt_sampling']
        NUM_LAST_EPOCHS: 4  # yield similar results if set it to 5 following mmdetection3d