Open wufeim opened 1 year ago
LittlePey
commented
1 year ago Hi, sorry for the late reply, the released model is for cars only and there is no plan to release the model for ped and cyc. You can modify our yaml file according to pv_rcnn.yaml. We did not do special design for ped and cyc when we train our 3 classes model.
wufeim
commented
1 year ago Sounds good. Thanks for the information.
CaptainXX
commented
1 year ago Sounds good. Thanks for the information.
Hello! Have you successfully trained the model for all classes? I changed the YAML file as the following below, and encountered an runtime error when training: RuntimeError: shape '[1, 800, 704, -1]' is invalid for input of size 70400 in function assign_target
file pcdet/models/dense_heads/target_assigner/axis_aligned_target_assigner.py.
CLASS_NAMES: ['Car', 'Pedestrian', 'Cyclist']
DATA_CONFIG:
_BASE_CONFIG_: cfgs/dataset_configs/kitti_dataset.yaml
DATASET: 'KittiDatasetSFD'
DATA_PATH: '../data/kitti_sfd_seguv_twise'
USE_VAN: True
DATA_AUGMENTOR:
DISABLE_AUG_LIST: ['placeholder']
AUG_CONFIG_LIST:
- NAME: gt_sampling_sfd
USE_ROAD_PLANE: FALSE
DB_INFO_PATH:
- kitti_dbinfos_train_sfd_seguv.pkl
PREPARE: {
filter_by_min_points: ['Car:5', 'Pedestrian:5', 'Cyclist:5'],
filter_by_difficulty: [-1],
}
SAMPLE_GROUPS: ['Car:15', 'Pedestrian:15', 'Cyclist:15']
NUM_POINT_FEATURES: 4
NUM_POINT_FEATURES_PSEUDO: 9
DATABASE_WITH_FAKELIDAR: False
REMOVE_EXTRA_WIDTH: [0.0, 0.0, 0.0]
LIMIT_WHOLE_SCENE: False
- NAME: random_local_noise_sfd
LOCAL_ROT_RANGE: [-0.78539816, 0.78539816]
TRANSLATION_STD: [1.0, 1.0, 0.5]
GLOBAL_ROT_RANGE: [0.0, 0.0]
EXTRA_WIDTH: [0.2, 0.2, 0.0]
- NAME: random_world_flip_sfd
ALONG_AXIS_LIST: ['x']
- NAME: random_world_rotation_sfd
WORLD_ROT_ANGLE: [-0.78539816, 0.78539816]
- NAME: random_world_scaling_sfd
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'],
encoding_type_pseudo: absolute_coordinates_encoding_pseudo,
used_feature_list_pseudo: ['x', 'y', 'z', 'r', 'g', 'b','u','v'],
src_feature_list_pseudo: ['x', 'y', 'z', 'r', 'g', 'b', 'mask', 'u','v'],
}
DATA_PROCESSOR:
- NAME: mask_points_and_boxes_outside_range_sfd
REMOVE_OUTSIDE_BOXES: True
- NAME: shuffle_points_sfd
SHUFFLE_ENABLED: {
'train': True,
'test': False
}
- NAME: transform_points_to_voxels_valid
VOXEL_SIZE: [0.05, 0.05, 0.1]
MAX_POINTS_PER_VOXEL: 5
MAX_NUMBER_OF_VOXELS: {
'train': 16000,
'test': 40000
}
- NAME: grid_sample_points_pseudo
MAX_DISTANCE: 13
MODEL:
NAME: SFD
VFE:
NAME: MeanVFE
BACKBONE_3D:
NAME: VoxelBackBone8x
MAP_TO_BEV:
NAME: HeightCompression
NUM_BEV_FEATURES: 256
BACKBONE_2D:
NAME: BaseBEVBackbone
LAYER_NUMS: [4, 4]
LAYER_STRIDES: [1, 2]
NUM_FILTERS: [64, 128]
UPSAMPLE_STRIDES: [1, 2]
NUM_UPSAMPLE_FILTERS: [128, 128]
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: [
{
'class_name': 'Car',
'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,
'matched_threshold': 0.6,
'unmatched_threshold': 0.45
},
{
'class_name': 'Pedestrian',
'anchor_sizes': [[0.8, 0.6, 1.73]],
'anchor_rotations': [0, 1.57],
'anchor_bottom_heights': [-0.6],
'align_center': False,
'feature_map_stride': 2,
'matched_threshold': 0.5,
'unmatched_threshold': 0.35
},
{
'class_name': 'Cyclist',
'anchor_sizes': [[1.76, 0.6, 1.73]],
'anchor_rotations': [0, 1.57],
'anchor_bottom_heights': [-0.6],
'align_center': False,
'feature_map_stride': 2,
'matched_threshold': 0.5,
'unmatched_threshold': 0.35
}
]
TARGET_ASSIGNER_CONFIG:
NAME: AxisAlignedTargetAssigner
POS_FRACTION: -1.0
SAMPLE_SIZE: 512
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]
}
ROI_HEAD:
NAME: SFDHead
CLASS_AGNOSTIC: True
ATTENTION_FC: [256]
SHARED_FC: [512,512]
CLS_FC: [512, 512]
REG_FC: [512, 512]
DP_RATIO: 0.3
SHARED_FC_PSEUDO: [256, 256]
CLS_FC_PSEUDO: [256, 256]
REG_FC_PSEUDO: [256, 256]
SHARED_FC_VALID: [256, 256]
CLS_FC_VALID: [256, 256]
REG_FC_VALID: [256, 256]
AUXILIARY_CODE_SIZE: 7
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
USE_FAST_NMS: True
SCORE_THRESH: 0.0
NMS_PRE_MAXSIZE: 2048
NMS_POST_MAXSIZE: 100
NMS_THRESH: 0.7
ROI_GRID_POOL:
FEATURES_SOURCE: ['x_conv3', 'x_conv4']
PRE_MLP: True
GRID_SIZE: 6
POOL_LAYERS:
x_conv3:
MLPS: [[32, 32], [32, 32]]
QUERY_RANGES: [[2, 2, 2], [4, 4, 4]]
POOL_RADIUS: [0.4, 0.8]
NSAMPLE: [16, 16]
POOL_METHOD: max_pool
x_conv4:
MLPS: [[32, 32], [32, 32]]
QUERY_RANGES: [[2, 2, 2], [4, 4, 4]]
POOL_RADIUS: [0.8, 1.6]
NSAMPLE: [16, 16]
POOL_METHOD: max_pool
ROI_POINT_CROP:
POOL_EXTRA_WIDTH: [0.2,0.2,0.6]
DEPTH_NORMALIZER: 70.0
ROI_AWARE_POOL:
POOL_SIZE: 12
NUM_FEATURES: 128
MAX_POINTS_PER_VOXEL: 64
NUM_FEATURES_RAW: 90
POOL_METHOD: max
KERNEL_TYPE: coords_5x5_dilate
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: giou
REG_LOSS_PSEUDO: smooth-l1
REG_LOSS_VALID: smooth-l1
CORNER_LOSS_REGULARIZATION: True
GRID_3D_IOU_LOSS: False
LOSS_WEIGHTS: {
'rcnn_cls_weight': 1.0,
'rcnn_reg_weight': 1.0,
'rcnn_corner_weight': 1.0,
'rcnn_iou3d_weight': 1.0,
'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
}
LOSS_WEIGHTS_PSEUDO: {
'rcnn_cls_weight': 0.5,
'rcnn_reg_weight': 0.5,
'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
}
LOSS_WEIGHTS_VALID: {
'rcnn_cls_weight': 0.5,
'rcnn_reg_weight': 0.5,
'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: 0.3
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:
BATCH_SIZE_PER_GPU: 1
NUM_EPOCHS: 40
OPTIMIZER: adam_onecycle
LR: 0.001
WEIGHT_DECAY: 0.001
MOMENTUM: 0.09
MOMS: [0.95, 0.85]
PCT_START: 0.4
DIV_FACTOR: 10
DECAY_STEP_LIST: [15,25]
LR_DECAY: 0.1
LR_CLIP: 0.0000001
LR_WARMUP: False
WARMUP_EPOCH: 1
GRAD_NORM_CLIP: 10
Sounds good. Thanks for the information.
Hello! Have you successfully trained the model for all classes? I changed the YAML file as the following below, and encountered an runtime error when training:
RuntimeError: shape '[1, 800, 704, -1]' is invalid for input of size 70400in functionassign_targetfilepcdet/models/dense_heads/target_assigner/axis_aligned_target_assigner.py.CLASS_NAMES: ['Car', 'Pedestrian', 'Cyclist'] DATA_CONFIG: _BASE_CONFIG_: cfgs/dataset_configs/kitti_dataset.yaml DATASET: 'KittiDatasetSFD' DATA_PATH: '../data/kitti_sfd_seguv_twise' USE_VAN: True DATA_AUGMENTOR: DISABLE_AUG_LIST: ['placeholder'] AUG_CONFIG_LIST: - NAME: gt_sampling_sfd USE_ROAD_PLANE: FALSE DB_INFO_PATH: - kitti_dbinfos_train_sfd_seguv.pkl PREPARE: { filter_by_min_points: ['Car:5', 'Pedestrian:5', 'Cyclist:5'], filter_by_difficulty: [-1], } SAMPLE_GROUPS: ['Car:15', 'Pedestrian:15', 'Cyclist:15'] NUM_POINT_FEATURES: 4 NUM_POINT_FEATURES_PSEUDO: 9 DATABASE_WITH_FAKELIDAR: False REMOVE_EXTRA_WIDTH: [0.0, 0.0, 0.0] LIMIT_WHOLE_SCENE: False - NAME: random_local_noise_sfd LOCAL_ROT_RANGE: [-0.78539816, 0.78539816] TRANSLATION_STD: [1.0, 1.0, 0.5] GLOBAL_ROT_RANGE: [0.0, 0.0] EXTRA_WIDTH: [0.2, 0.2, 0.0] - NAME: random_world_flip_sfd ALONG_AXIS_LIST: ['x'] - NAME: random_world_rotation_sfd WORLD_ROT_ANGLE: [-0.78539816, 0.78539816] - NAME: random_world_scaling_sfd 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'], encoding_type_pseudo: absolute_coordinates_encoding_pseudo, used_feature_list_pseudo: ['x', 'y', 'z', 'r', 'g', 'b','u','v'], src_feature_list_pseudo: ['x', 'y', 'z', 'r', 'g', 'b', 'mask', 'u','v'], } DATA_PROCESSOR: - NAME: mask_points_and_boxes_outside_range_sfd REMOVE_OUTSIDE_BOXES: True - NAME: shuffle_points_sfd SHUFFLE_ENABLED: { 'train': True, 'test': False } - NAME: transform_points_to_voxels_valid VOXEL_SIZE: [0.05, 0.05, 0.1] MAX_POINTS_PER_VOXEL: 5 MAX_NUMBER_OF_VOXELS: { 'train': 16000, 'test': 40000 } - NAME: grid_sample_points_pseudo MAX_DISTANCE: 13 MODEL: NAME: SFD VFE: NAME: MeanVFE BACKBONE_3D: NAME: VoxelBackBone8x MAP_TO_BEV: NAME: HeightCompression NUM_BEV_FEATURES: 256 BACKBONE_2D: NAME: BaseBEVBackbone LAYER_NUMS: [4, 4] LAYER_STRIDES: [1, 2] NUM_FILTERS: [64, 128] UPSAMPLE_STRIDES: [1, 2] NUM_UPSAMPLE_FILTERS: [128, 128] 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: [ { 'class_name': 'Car', '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, 'matched_threshold': 0.6, 'unmatched_threshold': 0.45 }, { 'class_name': 'Pedestrian', 'anchor_sizes': [[0.8, 0.6, 1.73]], 'anchor_rotations': [0, 1.57], 'anchor_bottom_heights': [-0.6], 'align_center': False, 'feature_map_stride': 2, 'matched_threshold': 0.5, 'unmatched_threshold': 0.35 }, { 'class_name': 'Cyclist', 'anchor_sizes': [[1.76, 0.6, 1.73]], 'anchor_rotations': [0, 1.57], 'anchor_bottom_heights': [-0.6], 'align_center': False, 'feature_map_stride': 2, 'matched_threshold': 0.5, 'unmatched_threshold': 0.35 } ] TARGET_ASSIGNER_CONFIG: NAME: AxisAlignedTargetAssigner POS_FRACTION: -1.0 SAMPLE_SIZE: 512 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] } ROI_HEAD: NAME: SFDHead CLASS_AGNOSTIC: True ATTENTION_FC: [256] SHARED_FC: [512,512] CLS_FC: [512, 512] REG_FC: [512, 512] DP_RATIO: 0.3 SHARED_FC_PSEUDO: [256, 256] CLS_FC_PSEUDO: [256, 256] REG_FC_PSEUDO: [256, 256] SHARED_FC_VALID: [256, 256] CLS_FC_VALID: [256, 256] REG_FC_VALID: [256, 256] AUXILIARY_CODE_SIZE: 7 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 USE_FAST_NMS: True SCORE_THRESH: 0.0 NMS_PRE_MAXSIZE: 2048 NMS_POST_MAXSIZE: 100 NMS_THRESH: 0.7 ROI_GRID_POOL: FEATURES_SOURCE: ['x_conv3', 'x_conv4'] PRE_MLP: True GRID_SIZE: 6 POOL_LAYERS: x_conv3: MLPS: [[32, 32], [32, 32]] QUERY_RANGES: [[2, 2, 2], [4, 4, 4]] POOL_RADIUS: [0.4, 0.8] NSAMPLE: [16, 16] POOL_METHOD: max_pool x_conv4: MLPS: [[32, 32], [32, 32]] QUERY_RANGES: [[2, 2, 2], [4, 4, 4]] POOL_RADIUS: [0.8, 1.6] NSAMPLE: [16, 16] POOL_METHOD: max_pool ROI_POINT_CROP: POOL_EXTRA_WIDTH: [0.2,0.2,0.6] DEPTH_NORMALIZER: 70.0 ROI_AWARE_POOL: POOL_SIZE: 12 NUM_FEATURES: 128 MAX_POINTS_PER_VOXEL: 64 NUM_FEATURES_RAW: 90 POOL_METHOD: max KERNEL_TYPE: coords_5x5_dilate 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: giou REG_LOSS_PSEUDO: smooth-l1 REG_LOSS_VALID: smooth-l1 CORNER_LOSS_REGULARIZATION: True GRID_3D_IOU_LOSS: False LOSS_WEIGHTS: { 'rcnn_cls_weight': 1.0, 'rcnn_reg_weight': 1.0, 'rcnn_corner_weight': 1.0, 'rcnn_iou3d_weight': 1.0, 'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0] } LOSS_WEIGHTS_PSEUDO: { 'rcnn_cls_weight': 0.5, 'rcnn_reg_weight': 0.5, 'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0] } LOSS_WEIGHTS_VALID: { 'rcnn_cls_weight': 0.5, 'rcnn_reg_weight': 0.5, '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: 0.3 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: BATCH_SIZE_PER_GPU: 1 NUM_EPOCHS: 40 OPTIMIZER: adam_onecycle LR: 0.001 WEIGHT_DECAY: 0.001 MOMENTUM: 0.09 MOMS: [0.95, 0.85] PCT_START: 0.4 DIV_FACTOR: 10 DECAY_STEP_LIST: [15,25] LR_DECAY: 0.1 LR_CLIP: 0.0000001 LR_WARMUP: False WARMUP_EPOCH: 1 GRAD_NORM_CLIP: 10
No I haven't spent time on this yet.
HuangLLL123
commented
7 months ago Sounds good. Thanks for the information.听起来不错谢谢你的信息。
Hello! Have you successfully trained the model for all classes? I changed the YAML file as the following below, and encountered an runtime error when training:
RuntimeError: shape '[1, 800, 704, -1]' is invalid for input of size 70400in functionassign_target你好啊！你是否成功地训练了所有类的模型？我将YAML文件更改为以下内容，并在训练时遇到运行时错误：RuntimeError: shape '[1, 800, 704, -1]' is invalid for input of size 70400in functionassign_targetfilepcdet/models/dense_heads/target_assigner/axis_aligned_target_assigner.py. 文件pcdet/models/dense_heads/target_assigner/axis_aligned_target_assigner.py。CLASS_NAMES: ['Car', 'Pedestrian', 'Cyclist'] DATA_CONFIG: _BASE_CONFIG_: cfgs/dataset_configs/kitti_dataset.yaml DATASET: 'KittiDatasetSFD' DATA_PATH: '../data/kitti_sfd_seguv_twise' USE_VAN: True DATA_AUGMENTOR: DISABLE_AUG_LIST: ['placeholder'] AUG_CONFIG_LIST: - NAME: gt_sampling_sfd USE_ROAD_PLANE: FALSE DB_INFO_PATH: - kitti_dbinfos_train_sfd_seguv.pkl PREPARE: { filter_by_min_points: ['Car:5', 'Pedestrian:5', 'Cyclist:5'], filter_by_difficulty: [-1], } SAMPLE_GROUPS: ['Car:15', 'Pedestrian:15', 'Cyclist:15'] NUM_POINT_FEATURES: 4 NUM_POINT_FEATURES_PSEUDO: 9 DATABASE_WITH_FAKELIDAR: False REMOVE_EXTRA_WIDTH: [0.0, 0.0, 0.0] LIMIT_WHOLE_SCENE: False - NAME: random_local_noise_sfd LOCAL_ROT_RANGE: [-0.78539816, 0.78539816] TRANSLATION_STD: [1.0, 1.0, 0.5] GLOBAL_ROT_RANGE: [0.0, 0.0] EXTRA_WIDTH: [0.2, 0.2, 0.0] - NAME: random_world_flip_sfd ALONG_AXIS_LIST: ['x'] - NAME: random_world_rotation_sfd WORLD_ROT_ANGLE: [-0.78539816, 0.78539816] - NAME: random_world_scaling_sfd 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'], encoding_type_pseudo: absolute_coordinates_encoding_pseudo, used_feature_list_pseudo: ['x', 'y', 'z', 'r', 'g', 'b','u','v'], src_feature_list_pseudo: ['x', 'y', 'z', 'r', 'g', 'b', 'mask', 'u','v'], } DATA_PROCESSOR: - NAME: mask_points_and_boxes_outside_range_sfd REMOVE_OUTSIDE_BOXES: True - NAME: shuffle_points_sfd SHUFFLE_ENABLED: { 'train': True, 'test': False } - NAME: transform_points_to_voxels_valid VOXEL_SIZE: [0.05, 0.05, 0.1] MAX_POINTS_PER_VOXEL: 5 MAX_NUMBER_OF_VOXELS: { 'train': 16000, 'test': 40000 } - NAME: grid_sample_points_pseudo MAX_DISTANCE: 13 MODEL: NAME: SFD VFE: NAME: MeanVFE BACKBONE_3D: NAME: VoxelBackBone8x MAP_TO_BEV: NAME: HeightCompression NUM_BEV_FEATURES: 256 BACKBONE_2D: NAME: BaseBEVBackbone LAYER_NUMS: [4, 4] LAYER_STRIDES: [1, 2] NUM_FILTERS: [64, 128] UPSAMPLE_STRIDES: [1, 2] NUM_UPSAMPLE_FILTERS: [128, 128] 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: [ { 'class_name': 'Car', '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, 'matched_threshold': 0.6, 'unmatched_threshold': 0.45 }, { 'class_name': 'Pedestrian', 'anchor_sizes': [[0.8, 0.6, 1.73]], 'anchor_rotations': [0, 1.57], 'anchor_bottom_heights': [-0.6], 'align_center': False, 'feature_map_stride': 2, 'matched_threshold': 0.5, 'unmatched_threshold': 0.35 }, { 'class_name': 'Cyclist', 'anchor_sizes': [[1.76, 0.6, 1.73]], 'anchor_rotations': [0, 1.57], 'anchor_bottom_heights': [-0.6], 'align_center': False, 'feature_map_stride': 2, 'matched_threshold': 0.5, 'unmatched_threshold': 0.35 } ] TARGET_ASSIGNER_CONFIG: NAME: AxisAlignedTargetAssigner POS_FRACTION: -1.0 SAMPLE_SIZE: 512 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] } ROI_HEAD: NAME: SFDHead CLASS_AGNOSTIC: True ATTENTION_FC: [256] SHARED_FC: [512,512] CLS_FC: [512, 512] REG_FC: [512, 512] DP_RATIO: 0.3 SHARED_FC_PSEUDO: [256, 256] CLS_FC_PSEUDO: [256, 256] REG_FC_PSEUDO: [256, 256] SHARED_FC_VALID: [256, 256] CLS_FC_VALID: [256, 256] REG_FC_VALID: [256, 256] AUXILIARY_CODE_SIZE: 7 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 USE_FAST_NMS: True SCORE_THRESH: 0.0 NMS_PRE_MAXSIZE: 2048 NMS_POST_MAXSIZE: 100 NMS_THRESH: 0.7 ROI_GRID_POOL: FEATURES_SOURCE: ['x_conv3', 'x_conv4'] PRE_MLP: True GRID_SIZE: 6 POOL_LAYERS: x_conv3: MLPS: [[32, 32], [32, 32]] QUERY_RANGES: [[2, 2, 2], [4, 4, 4]] POOL_RADIUS: [0.4, 0.8] NSAMPLE: [16, 16] POOL_METHOD: max_pool x_conv4: MLPS: [[32, 32], [32, 32]] QUERY_RANGES: [[2, 2, 2], [4, 4, 4]] POOL_RADIUS: [0.8, 1.6] NSAMPLE: [16, 16] POOL_METHOD: max_pool ROI_POINT_CROP: POOL_EXTRA_WIDTH: [0.2,0.2,0.6] DEPTH_NORMALIZER: 70.0 ROI_AWARE_POOL: POOL_SIZE: 12 NUM_FEATURES: 128 MAX_POINTS_PER_VOXEL: 64 NUM_FEATURES_RAW: 90 POOL_METHOD: max KERNEL_TYPE: coords_5x5_dilate 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: giou REG_LOSS_PSEUDO: smooth-l1 REG_LOSS_VALID: smooth-l1 CORNER_LOSS_REGULARIZATION: True GRID_3D_IOU_LOSS: False LOSS_WEIGHTS: { 'rcnn_cls_weight': 1.0, 'rcnn_reg_weight': 1.0, 'rcnn_corner_weight': 1.0, 'rcnn_iou3d_weight': 1.0, 'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0] } LOSS_WEIGHTS_PSEUDO: { 'rcnn_cls_weight': 0.5, 'rcnn_reg_weight': 0.5, 'code_weights': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0] } LOSS_WEIGHTS_VALID: { 'rcnn_cls_weight': 0.5, 'rcnn_reg_weight': 0.5, '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: 0.3 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: BATCH_SIZE_PER_GPU: 1 NUM_EPOCHS: 40 OPTIMIZER: adam_onecycle LR: 0.001 WEIGHT_DECAY: 0.001 MOMENTUM: 0.09 MOMS: [0.95, 0.85] PCT_START: 0.4 DIV_FACTOR: 10 DECAY_STEP_LIST: [15,25] LR_DECAY: 0.1 LR_CLIP: 0.0000001 LR_WARMUP: False WARMUP_EPOCH: 1 GRAD_NORM_CLIP: 10
change the config file like this
{ 'class_name': 'Pedestrian', '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,
'feature_map_stride': 2,
'matched_threshold': 0.5, 'unmatched_threshold': 0.35 }, { 'class_name': 'Cyclist', '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,
'feature_map_stride': 2,
'matched_threshold': 0.5, 'unmatched_threshold': 0.35 }
Hi,
Thanks for releasing the code and trained models. I'm wondering if the released model is meant for cars only. If so, are there plans to release models for cyclists and pedestrians?
Thanks.