Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Only half of the inference results are displayed when run demo.py on the DENSE dataset #75

Open
RhythmOfTheRain-Byte opened this issue Dec 21, 2023 · 0 comments

Comments

@RhythmOfTheRain-Byte
Copy link

Excuse me, I am training the dense dataset using OpenPCDet. However, when I use the trained model's .pth file for inference and run demo.py, it only displays half of the inference results. Changing the range of points doesn't seem to resolve the issue either. Which is shown as fellows. Could you please help me understand the reason behind this?

Additionally, the results obtained during the evaluation phase are as follows.The discrepancies between bbox AP, bev AP, and 3D AP are quite significant. I am not sure if this is reasonable. Can someone help me understand this?
Car [email protected], 0.70, 0.70:
bbox AP:81.2543, 81.2543, 81.2543
bev AP:32.0643, 32.0643, 32.0643
3d AP:21.4249, 21.4249, 21.4249
aos AP:58.71, 58.71, 58.71
Car [email protected], 0.70, 0.70:
bbox AP:82.4550, 82.4550, 82.4550
bev AP:28.3860, 28.3860, 28.3860
3d AP:15.9582, 15.9582, 15.9582
aos AP:59.12, 59.12, 59.12
Car [email protected], 0.50, 0.50:
bbox AP:81.2543, 81.2543, 81.2543
bev AP:43.2742, 43.2742, 43.2742
3d AP:40.7021, 40.7021, 40.7021
aos AP:58.71, 58.71, 58.71
Car [email protected], 0.50, 0.50:
bbox AP:82.4550, 82.4550, 82.4550
bev AP:40.8015, 40.8015, 40.8015
3d AP:37.2809, 37.2809, 37.2809
aos AP:59.12, 59.12, 59.12

The model i choose is pillarnet and Part of the log file is as fellows:
2023-12-15 12:33:41,122 INFO cfg.CLASS_NAMES: ['Pedestrian', 'PassengerCar', 'RidableVehicle']
2023-12-15 12:33:41,122 INFO ----------- DATA_CONFIG -----------
2023-12-15 12:33:41,122 INFO cfg.DATA_CONFIG.DATASET: DenseDataset
2023-12-15 12:33:41,122 INFO cfg.DATA_CONFIG.DATA_PATH: data/dense
2023-12-15 12:33:41,122 INFO cfg.DATA_CONFIG.POINT_CLOUD_RANGE: [-70.4, -40, -3, 70.4, 40, 1]
2023-12-15 12:33:41,122 INFO ----------- MAP_CLASS_TO_KITTI -----------
2023-12-15 12:33:41,122 INFO cfg.DATA_CONFIG.MAP_CLASS_TO_KITTI.Pedestrian: Pedestrian
2023-12-15 12:33:41,122 INFO cfg.DATA_CONFIG.MAP_CLASS_TO_KITTI.PassengerCar: Car
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.MAP_CLASS_TO_KITTI.RidableVehicle: Cyclist
2023-12-15 12:33:41,123 INFO ----------- DATA_SPLIT -----------
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.DATA_SPLIT.train: train_clear
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.DATA_SPLIT.test: val_clear
2023-12-15 12:33:41,123 INFO ----------- INFO_PATH -----------
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.INFO_PATH.train: ['dense_infos_train_clear.pkl']
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.INFO_PATH.test: ['dense_infos_val_clear.pkl']
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.FOV_POINTS_ONLY: False
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.FOG_AUGMENTATION: False
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.SENSOR_TYPE: hdl64
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.SIGNAL_TYPE: strongest
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.COMPENSATE: [0.0, 0.0, 0.0]
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.STRONGEST_LAST_FILTER: False
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.DROP_EMPTY_ANNOTATIONS: True
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.FOG_AUGMENTATION_AFTER: False
2023-12-15 12:33:41,123 INFO ----------- DATA_AUGMENTOR -----------
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR.DISABLE_AUG_LIST: ['placeholder']
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR.AUG_CONFIG_LIST: [{'NAME': 'gt_sampling', 'USE_ROAD_PLANE': False, 'DB_INFO_PATH': ['dense_dbinfos_train_clear.pkl'], 'PREPARE': {'filter_by_min_points': ['Pedestrian:5', 'PassengerCar:5', 'RidableVehicle:5'], 'filter_by_difficulty': [-1]}, 'SAMPLE_GROUPS': ['Pedestrian:15', 'PassengerCar:20', 'RidableVehicle:15'], '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']}, {'NAME': 'random_world_rotation', 'WORLD_ROT_ANGLE': [-0.78539816, 0.78539816]}, {'NAME': 'random_world_scaling', 'WORLD_SCALE_RANGE': [0.95, 1.05]}]
2023-12-15 12:33:41,123 INFO ----------- POINT_FEATURE_ENCODING -----------
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.encoding_type: absolute_coordinates_encoding
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.used_feature_list: ['x', 'y', 'z', 'intensity']
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.src_feature_list: ['x', 'y', 'z', 'intensity', 'channel']
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.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}}]
2023-12-15 12:33:41,123 INFO cfg.DATA_CONFIG.BASE_CONFIG: tools/cfgs/dataset_configs/dense_dataset.yaml
Screenshot from 2023-12-19 13-10-43

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant