File size: 15,114 Bytes
4c8765e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 |
2023-04-01 21:45:09,865 INFO **********************Start logging********************** 2023-04-01 21:45:09,866 INFO CUDA_VISIBLE_DEVICES=ALL 2023-04-01 21:45:09,866 INFO total_batch_size: 16 2023-04-01 21:45:09,867 INFO cfg_file cfgs/scannet_models/CAGroup3D.yaml 2023-04-01 21:45:09,868 INFO batch_size 16 2023-04-01 21:45:09,868 INFO workers 4 2023-04-01 21:45:09,869 INFO extra_tag cagroup3d-win10-scannet-eval 2023-04-01 21:45:09,869 INFO ckpt ../output/scannet_models/CAGroup3D/cagroup3d-win10-scannet-train/ckpt/checkpoint_epoch_8.pth 2023-04-01 21:45:09,870 INFO launcher pytorch 2023-04-01 21:45:09,870 INFO tcp_port 18888 2023-04-01 21:45:09,870 INFO set_cfgs None 2023-04-01 21:45:09,872 INFO max_waiting_mins 30 2023-04-01 21:45:09,872 INFO start_epoch 0 2023-04-01 21:45:09,872 INFO eval_tag default 2023-04-01 21:45:09,872 INFO eval_all False 2023-04-01 21:45:09,873 INFO ckpt_dir None 2023-04-01 21:45:09,873 INFO save_to_file False 2023-04-01 21:45:09,875 INFO cfg.ROOT_DIR: C:\CITYU\CS5182\proj\CAGroup3D 2023-04-01 21:45:09,876 INFO cfg.LOCAL_RANK: 0 2023-04-01 21:45:09,876 INFO cfg.CLASS_NAMES: ['cabinet', 'bed', 'chair', 'sofa', 'table', 'door', 'window', 'bookshelf', 'picture', 'counter', 'desk', 'curtain', 'refrigerator', 'showercurtrain', 'toilet', 'sink', 'bathtub', 'garbagebin'] 2023-04-01 21:45:09,876 INFO cfg.DATA_CONFIG = edict() 2023-04-01 21:45:09,878 INFO cfg.DATA_CONFIG.DATASET: ScannetDataset 2023-04-01 21:45:09,878 INFO cfg.DATA_CONFIG.DATA_PATH: ../data/scannet_data/ScanNetV2 2023-04-01 21:45:09,879 INFO cfg.DATA_CONFIG.PROCESSED_DATA_TAG: scannet_processed_data_v0_5_0 2023-04-01 21:45:09,879 INFO cfg.DATA_CONFIG.POINT_CLOUD_RANGE: [-40, -40, -10, 40, 40, 10] 2023-04-01 21:45:09,880 INFO cfg.DATA_CONFIG.DATA_SPLIT = edict() 2023-04-01 21:45:09,880 INFO cfg.DATA_CONFIG.DATA_SPLIT.train: train 2023-04-01 21:45:09,880 INFO cfg.DATA_CONFIG.DATA_SPLIT.test: val 2023-04-01 21:45:09,882 INFO cfg.DATA_CONFIG.REPEAT = edict() 2023-04-01 21:45:09,883 INFO cfg.DATA_CONFIG.REPEAT.train: 10 2023-04-01 21:45:09,883 INFO cfg.DATA_CONFIG.REPEAT.test: 1 2023-04-01 21:45:09,884 INFO cfg.DATA_CONFIG.INFO_PATH = edict() 2023-04-01 21:45:09,884 INFO cfg.DATA_CONFIG.INFO_PATH.train: ['scannet_infos_train.pkl'] 2023-04-01 21:45:09,885 INFO cfg.DATA_CONFIG.INFO_PATH.test: ['scannet_infos_val.pkl'] 2023-04-01 21:45:09,886 INFO cfg.DATA_CONFIG.GET_ITEM_LIST: ['points', 'instance_mask', 'semantic_mask'] 2023-04-01 21:45:09,886 INFO cfg.DATA_CONFIG.FILTER_EMPTY_BOXES_FOR_TRAIN: True 2023-04-01 21:45:09,887 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR_TRAIN = edict() 2023-04-01 21:45:09,887 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR_TRAIN.DISABLE_AUG_LIST: ['placeholder'] 2023-04-01 21:45:09,888 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR_TRAIN.AUG_CONFIG_LIST: [{'NAME': 'global_alignment', 'rotation_axis': 2}, {'NAME': 'point_seg_class_mapping', 'valid_cat_ids': [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 16, 24, 28, 33, 34, 36, 39], 'max_cat_id': 40}, {'NAME': 'random_world_flip', 'ALONG_AXIS_LIST': ['x', 'y']}, {'NAME': 'random_world_rotation', 'WORLD_ROT_ANGLE': [-0.087266, 0.087266]}, {'NAME': 'random_world_scaling', 'WORLD_SCALE_RANGE': [0.9, 1.1]}, {'NAME': 'random_world_translation', 'ALONG_AXIS_LIST': ['x', 'y', 'z'], 'NOISE_TRANSLATE_STD': 0.1}] 2023-04-01 21:45:09,890 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR_TEST = edict() 2023-04-01 21:45:09,890 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR_TEST.DISABLE_AUG_LIST: ['placeholder'] 2023-04-01 21:45:09,890 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR_TEST.AUG_CONFIG_LIST: [{'NAME': 'global_alignment', 'rotation_axis': 2}] 2023-04-01 21:45:09,890 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR = edict() 2023-04-01 21:45:09,892 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR.DISABLE_AUG_LIST: ['placeholder'] 2023-04-01 21:45:09,893 INFO cfg.DATA_CONFIG.DATA_AUGMENTOR.AUG_CONFIG_LIST: [{'NAME': 'global_alignment', 'rotation_axis': 2}] 2023-04-01 21:45:09,893 INFO cfg.DATA_CONFIG.POINT_FEATURE_ENCODING = edict() 2023-04-01 21:45:09,894 INFO cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.encoding_type: absolute_coordinates_encoding 2023-04-01 21:45:09,895 INFO cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.used_feature_list: ['x', 'y', 'z', 'r', 'g', 'b'] 2023-04-01 21:45:09,895 INFO cfg.DATA_CONFIG.POINT_FEATURE_ENCODING.src_feature_list: ['x', 'y', 'z', 'r', 'g', 'b'] 2023-04-01 21:45:09,896 INFO cfg.DATA_CONFIG.DATA_PROCESSOR: [{'NAME': 'mask_points_and_boxes_outside_range', 'REMOVE_OUTSIDE_BOXES': True}] 2023-04-01 21:45:09,896 INFO cfg.DATA_CONFIG._BASE_CONFIG_: cfgs/dataset_configs/scannet_dataset.yaml 2023-04-01 21:45:09,896 INFO cfg.VOXEL_SIZE: 0.02 2023-04-01 21:45:09,896 INFO cfg.N_CLASSES: 18 2023-04-01 21:45:09,898 INFO cfg.SEMANTIC_THR: 0.15 2023-04-01 21:45:09,898 INFO cfg.MODEL = edict() 2023-04-01 21:45:09,899 INFO cfg.MODEL.NAME: CAGroup3D 2023-04-01 21:45:09,899 INFO cfg.MODEL.VOXEL_SIZE: 0.02 2023-04-01 21:45:09,900 INFO cfg.MODEL.SEMANTIC_MIN_THR: 0.05 2023-04-01 21:45:09,900 INFO cfg.MODEL.SEMANTIC_ITER_VALUE: 0.02 2023-04-01 21:45:09,900 INFO cfg.MODEL.SEMANTIC_THR: 0.15 2023-04-01 21:45:09,901 INFO cfg.MODEL.BACKBONE_3D = edict() 2023-04-01 21:45:09,901 INFO cfg.MODEL.BACKBONE_3D.NAME: BiResNet 2023-04-01 21:45:09,902 INFO cfg.MODEL.BACKBONE_3D.IN_CHANNELS: 3 2023-04-01 21:45:09,902 INFO cfg.MODEL.BACKBONE_3D.OUT_CHANNELS: 64 2023-04-01 21:45:09,902 INFO cfg.MODEL.DENSE_HEAD = edict() 2023-04-01 21:45:09,903 INFO cfg.MODEL.DENSE_HEAD.NAME: CAGroup3DHead 2023-04-01 21:45:09,903 INFO cfg.MODEL.DENSE_HEAD.IN_CHANNELS: [64, 128, 256, 512] 2023-04-01 21:45:09,904 INFO cfg.MODEL.DENSE_HEAD.OUT_CHANNELS: 64 2023-04-01 21:45:09,904 INFO cfg.MODEL.DENSE_HEAD.SEMANTIC_THR: 0.15 2023-04-01 21:45:09,905 INFO cfg.MODEL.DENSE_HEAD.VOXEL_SIZE: 0.02 2023-04-01 21:45:09,905 INFO cfg.MODEL.DENSE_HEAD.N_CLASSES: 18 2023-04-01 21:45:09,906 INFO cfg.MODEL.DENSE_HEAD.N_REG_OUTS: 6 2023-04-01 21:45:09,906 INFO cfg.MODEL.DENSE_HEAD.CLS_KERNEL: 9 2023-04-01 21:45:09,906 INFO cfg.MODEL.DENSE_HEAD.WITH_YAW: False 2023-04-01 21:45:09,907 INFO cfg.MODEL.DENSE_HEAD.USE_SEM_SCORE: False 2023-04-01 21:45:09,907 INFO cfg.MODEL.DENSE_HEAD.EXPAND_RATIO: 3 2023-04-01 21:45:09,908 INFO cfg.MODEL.DENSE_HEAD.ASSIGNER = edict() 2023-04-01 21:45:09,908 INFO cfg.MODEL.DENSE_HEAD.ASSIGNER.NAME: CAGroup3DAssigner 2023-04-01 21:45:09,909 INFO cfg.MODEL.DENSE_HEAD.ASSIGNER.LIMIT: 27 2023-04-01 21:45:09,909 INFO cfg.MODEL.DENSE_HEAD.ASSIGNER.TOPK: 18 2023-04-01 21:45:09,909 INFO cfg.MODEL.DENSE_HEAD.ASSIGNER.N_SCALES: 4 2023-04-01 21:45:09,910 INFO cfg.MODEL.DENSE_HEAD.LOSS_OFFSET = edict() 2023-04-01 21:45:09,910 INFO cfg.MODEL.DENSE_HEAD.LOSS_OFFSET.NAME: SmoothL1Loss 2023-04-01 21:45:09,911 INFO cfg.MODEL.DENSE_HEAD.LOSS_OFFSET.BETA: 0.04 2023-04-01 21:45:09,911 INFO cfg.MODEL.DENSE_HEAD.LOSS_OFFSET.REDUCTION: sum 2023-04-01 21:45:09,912 INFO cfg.MODEL.DENSE_HEAD.LOSS_OFFSET.LOSS_WEIGHT: 1.0 2023-04-01 21:45:09,912 INFO cfg.MODEL.DENSE_HEAD.LOSS_BBOX = edict() 2023-04-01 21:45:09,913 INFO cfg.MODEL.DENSE_HEAD.LOSS_BBOX.NAME: IoU3DLoss 2023-04-01 21:45:09,913 INFO cfg.MODEL.DENSE_HEAD.LOSS_BBOX.WITH_YAW: False 2023-04-01 21:45:09,914 INFO cfg.MODEL.DENSE_HEAD.LOSS_BBOX.LOSS_WEIGHT: 1.0 2023-04-01 21:45:09,914 INFO cfg.MODEL.DENSE_HEAD.NMS_CONFIG = edict() 2023-04-01 21:45:09,915 INFO cfg.MODEL.DENSE_HEAD.NMS_CONFIG.SCORE_THR: 0.01 2023-04-01 21:45:09,915 INFO cfg.MODEL.DENSE_HEAD.NMS_CONFIG.NMS_PRE: 1000 2023-04-01 21:45:09,916 INFO cfg.MODEL.DENSE_HEAD.NMS_CONFIG.IOU_THR: 0.5 2023-04-01 21:45:09,916 INFO cfg.MODEL.ROI_HEAD = edict() 2023-04-01 21:45:09,917 INFO cfg.MODEL.ROI_HEAD.NAME: CAGroup3DRoIHead 2023-04-01 21:45:09,917 INFO cfg.MODEL.ROI_HEAD.NUM_CLASSES: 18 2023-04-01 21:45:09,918 INFO cfg.MODEL.ROI_HEAD.MIDDLE_FEATURE_SOURCE: [3] 2023-04-01 21:45:09,918 INFO cfg.MODEL.ROI_HEAD.GRID_SIZE: 7 2023-04-01 21:45:09,919 INFO cfg.MODEL.ROI_HEAD.VOXEL_SIZE: 0.02 2023-04-01 21:45:09,919 INFO cfg.MODEL.ROI_HEAD.COORD_KEY: 2 2023-04-01 21:45:09,919 INFO cfg.MODEL.ROI_HEAD.MLPS: [[64, 128, 128]] 2023-04-01 21:45:09,920 INFO cfg.MODEL.ROI_HEAD.CODE_SIZE: 6 2023-04-01 21:45:09,920 INFO cfg.MODEL.ROI_HEAD.ENCODE_SINCOS: False 2023-04-01 21:45:09,921 INFO cfg.MODEL.ROI_HEAD.ROI_PER_IMAGE: 128 2023-04-01 21:45:09,921 INFO cfg.MODEL.ROI_HEAD.ROI_FG_RATIO: 0.9 2023-04-01 21:45:09,921 INFO cfg.MODEL.ROI_HEAD.REG_FG_THRESH: 0.3 2023-04-01 21:45:09,922 INFO cfg.MODEL.ROI_HEAD.ROI_CONV_KERNEL: 5 2023-04-01 21:45:09,922 INFO cfg.MODEL.ROI_HEAD.ENLARGE_RATIO: False 2023-04-01 21:45:09,923 INFO cfg.MODEL.ROI_HEAD.USE_IOU_LOSS: False 2023-04-01 21:45:09,923 INFO cfg.MODEL.ROI_HEAD.USE_GRID_OFFSET: False 2023-04-01 21:45:09,923 INFO cfg.MODEL.ROI_HEAD.USE_SIMPLE_POOLING: True 2023-04-01 21:45:09,923 INFO cfg.MODEL.ROI_HEAD.USE_CENTER_POOLING: True 2023-04-01 21:45:09,924 INFO cfg.MODEL.ROI_HEAD.LOSS_WEIGHTS = edict() 2023-04-01 21:45:09,925 INFO cfg.MODEL.ROI_HEAD.LOSS_WEIGHTS.RCNN_CLS_WEIGHT: 1.0 2023-04-01 21:45:09,927 INFO cfg.MODEL.ROI_HEAD.LOSS_WEIGHTS.RCNN_REG_WEIGHT: 1.0 2023-04-01 21:45:09,927 INFO cfg.MODEL.ROI_HEAD.LOSS_WEIGHTS.RCNN_IOU_WEIGHT: 1.0 2023-04-01 21:45:09,928 INFO cfg.MODEL.ROI_HEAD.LOSS_WEIGHTS.CODE_WEIGHT: [1.0, 1.0, 1.0, 1.0, 1.0, 1.0] 2023-04-01 21:45:09,928 INFO cfg.MODEL.POST_PROCESSING = edict() 2023-04-01 21:45:09,928 INFO cfg.MODEL.POST_PROCESSING.RECALL_THRESH_LIST: [0.25, 0.5] 2023-04-01 21:45:09,929 INFO cfg.MODEL.POST_PROCESSING.EVAL_METRIC: scannet 2023-04-01 21:45:09,929 INFO cfg.OPTIMIZATION = edict() 2023-04-01 21:45:09,931 INFO cfg.OPTIMIZATION.BATCH_SIZE_PER_GPU: 16 2023-04-01 21:45:09,931 INFO cfg.OPTIMIZATION.NUM_EPOCHS: 1 2023-04-01 21:45:09,931 INFO cfg.OPTIMIZATION.OPTIMIZER: adamW 2023-04-01 21:45:09,932 INFO cfg.OPTIMIZATION.LR: 0.001 2023-04-01 21:45:09,932 INFO cfg.OPTIMIZATION.WEIGHT_DECAY: 0.0001 2023-04-01 21:45:09,933 INFO cfg.OPTIMIZATION.DECAY_STEP_LIST: [7, 9] 2023-04-01 21:45:09,933 INFO cfg.OPTIMIZATION.LR_DECAY: 0.1 2023-04-01 21:45:09,933 INFO cfg.OPTIMIZATION.GRAD_NORM_CLIP: 10 2023-04-01 21:45:09,934 INFO cfg.OPTIMIZATION.PCT_START: 0.4 2023-04-01 21:45:09,934 INFO cfg.OPTIMIZATION.DIV_FACTOR: 10 2023-04-01 21:45:09,935 INFO cfg.OPTIMIZATION.LR_CLIP: 1e-07 2023-04-01 21:45:09,935 INFO cfg.OPTIMIZATION.LR_WARMUP: False 2023-04-01 21:45:09,936 INFO cfg.OPTIMIZATION.WARMUP_EPOCH: 1 2023-04-01 21:45:09,936 INFO cfg.TAG: CAGroup3D 2023-04-01 21:45:09,937 INFO cfg.EXP_GROUP_PATH: scannet_models 2023-04-01 21:45:09,937 INFO Loading SCANNET dataset 2023-04-01 21:45:09,972 INFO Total samples for SCANNET dataset: 312 2023-04-01 21:45:12,816 INFO ==> Loading parameters from checkpoint ../output/scannet_models/CAGroup3D/cagroup3d-win10-scannet-train/ckpt/checkpoint_epoch_8.pth to CPU 2023-04-01 21:45:13,846 INFO ==> Checkpoint trained from version: pcdet+0.5.2+0000000 2023-04-01 21:45:13,984 INFO ==> Done (loaded 838/838) 2023-04-01 21:45:14,393 INFO *************** EPOCH 8 EVALUATION ***************** 2023-04-02 00:01:12,334 INFO *************** Performance of EPOCH 8 ***************** 2023-04-02 00:01:12,334 INFO Generate label finished(sec_per_example: 26.1406 second). 2023-04-02 00:01:12,335 INFO recall_roi_0.25: 0.000000 2023-04-02 00:01:12,336 INFO recall_rcnn_0.25: 0.000000 2023-04-02 00:01:12,336 INFO recall_roi_0.5: 0.000000 2023-04-02 00:01:12,337 INFO recall_rcnn_0.5: 0.000000 2023-04-02 00:01:12,337 INFO Average predicted number of objects(312 samples): 490.029 2023-04-02 00:01:30,129 INFO {'cabinet_AP_0.25': 0.5813855528831482, 'bed_AP_0.25': 0.8816422820091248, 'chair_AP_0.25': 0.9543977379798889, 'sofa_AP_0.25': 0.8942864537239075, 'table_AP_0.25': 0.6969487071037292, 'door_AP_0.25': 0.682966947555542, 'window_AP_0.25': 0.6282612085342407, 'bookshelf_AP_0.25': 0.7065274715423584, 'picture_AP_0.25': 0.3937835395336151, 'counter_AP_0.25': 0.7749427556991577, 'desk_AP_0.25': 0.84075927734375, 'curtain_AP_0.25': 0.7117773294448853, 'refrigerator_AP_0.25': 0.5583232641220093, 'showercurtrain_AP_0.25': 0.7335410118103027, 'toilet_AP_0.25': 1.0, 'sink_AP_0.25': 0.7548432350158691, 'bathtub_AP_0.25': 0.87208491563797, 'garbagebin_AP_0.25': 0.6607850790023804, 'mAP_0.25': 0.7404030561447144, 'cabinet_rec_0.25': 0.9112903225806451, 'bed_rec_0.25': 0.9259259259259259, 'chair_rec_0.25': 0.9714912280701754, 'sofa_rec_0.25': 0.9690721649484536, 'table_rec_0.25': 0.8485714285714285, 'door_rec_0.25': 0.9036402569593148, 'window_rec_0.25': 0.875886524822695, 'bookshelf_rec_0.25': 0.9090909090909091, 'picture_rec_0.25': 0.6441441441441441, 'counter_rec_0.25': 0.9038461538461539, 'desk_rec_0.25': 0.968503937007874, 'curtain_rec_0.25': 0.8656716417910447, 'refrigerator_rec_0.25': 0.8421052631578947, 'showercurtrain_rec_0.25': 0.9642857142857143, 'toilet_rec_0.25': 1.0, 'sink_rec_0.25': 0.8469387755102041, 'bathtub_rec_0.25': 0.9032258064516129, 'garbagebin_rec_0.25': 0.8849056603773585, 'mAR_0.25': 0.8965886587523083, 'cabinet_AP_0.50': 0.4341818690299988, 'bed_AP_0.50': 0.8374742865562439, 'chair_AP_0.50': 0.9089729189872742, 'sofa_AP_0.50': 0.8061330914497375, 'table_AP_0.50': 0.6511608958244324, 'door_AP_0.50': 0.5602805614471436, 'window_AP_0.50': 0.3997173607349396, 'bookshelf_AP_0.50': 0.5989829897880554, 'picture_AP_0.50': 0.28941845893859863, 'counter_AP_0.50': 0.4558248519897461, 'desk_AP_0.50': 0.6606391072273254, 'curtain_AP_0.50': 0.5577377676963806, 'refrigerator_AP_0.50': 0.5126804113388062, 'showercurtrain_AP_0.50': 0.48787981271743774, 'toilet_AP_0.50': 0.958685576915741, 'sink_AP_0.50': 0.5293706655502319, 'bathtub_AP_0.50': 0.7828920483589172, 'garbagebin_AP_0.50': 0.5928370952606201, 'mAP_0.50': 0.6124927997589111, 'cabinet_rec_0.50': 0.7446236559139785, 'bed_rec_0.50': 0.8765432098765432, 'chair_rec_0.50': 0.9305555555555556, 'sofa_rec_0.50': 0.9072164948453608, 'table_rec_0.50': 0.7742857142857142, 'door_rec_0.50': 0.7601713062098501, 'window_rec_0.50': 0.648936170212766, 'bookshelf_rec_0.50': 0.7792207792207793, 'picture_rec_0.50': 0.481981981981982, 'counter_rec_0.50': 0.6153846153846154, 'desk_rec_0.50': 0.84251968503937, 'curtain_rec_0.50': 0.7313432835820896, 'refrigerator_rec_0.50': 0.7719298245614035, 'showercurtrain_rec_0.50': 0.6428571428571429, 'toilet_rec_0.50': 0.9655172413793104, 'sink_rec_0.50': 0.6224489795918368, 'bathtub_rec_0.50': 0.8387096774193549, 'garbagebin_rec_0.50': 0.7754716981132076, 'mAR_0.50': 0.7616509453350477} 2023-04-02 00:01:30,138 INFO Result is save to C:\CITYU\CS5182\proj\CAGroup3D\output\scannet_models\CAGroup3D\cagroup3d-win10-scannet-eval\eval\epoch_8\val\default 2023-04-02 00:01:30,139 INFO ****************Evaluation done.***************** |