search

traveller59 / second.pytorch

SECOND for KITTI/NuScenes object detection
MIT License
1.72k stars 722 forks source link

evaluation issue with nuscenes dataset #212

Open yalagamsrinivas opened 5 years ago

yalagamsrinivas commented 5 years ago

Generate output labels... [100.0%][===================>][4.49it/s][22:53>00:00]
generate label finished(4.38/s). start eval: Traceback (most recent call last): File "/media/5.5T_Harddisk/LIDAR/PointPillars_Nuscens/second.pytorch/second/data/nusc_eval.py", line 18, in fire.Fire(eval_main) File "/home/srinivas/.conda/envs/pointpillars/lib/python3.7/site-packages/fire/core.py", line 127, in Fire component_trace = _Fire(component, args, context, name) File "/home/srinivas/.conda/envs/pointpillars/lib/python3.7/site-packages/fire/core.py", line 366, in _Fire component, remaining_args) File "/home/srinivas/.conda/envs/pointpillars/lib/python3.7/site-packages/fire/core.py", line 542, in _CallCallable result = fn(*varargs, **kwargs) File "/media/5.5T_Harddisk/LIDAR/PointPillars_Nuscens/second.pytorch/second/data/nusc_eval.py", line 14, in eval_main verbose=False) File "/home/srinivas/.conda/envs/pointpillars/lib/python3.7/site-packages/nuscenes/eval/detection/evaluate.py", line 73, in init self.pred_boxes = load_prediction(self.result_path, self.cfg.max_boxes_per_sample, verbose=verbose) File "/home/srinivas/.conda/envs/pointpillars/lib/python3.7/site-packages/nuscenes/eval/detection/loaders.py", line 23, in load_prediction all_results = EvalBoxes.deserialize(json.load(f)) File "/home/srinivas/.conda/envs/pointpillars/lib/python3.7/site-packages/nuscenes/eval/detection/data_classes.py", line 216, in deserialize eb.add_boxes(sample_token, [EvalBox.deserialize(box) for box in boxes]) File "/home/srinivas/.conda/envs/pointpillars/lib/python3.7/site-packages/nuscenes/eval/detection/data_classes.py", line 216, in eb.add_boxes(sample_token, [EvalBox.deserialize(box) for box in boxes]) File "/home/srinivas/.conda/envs/pointpillars/lib/python3.7/site-packages/nuscenes/eval/detection/data_classes.py", line 155, in deserialize return cls(sample_token=content['sample_token'], TypeError: string indices must be integers

yalagamsrinivas commented 5 years ago

File "/home/srinivas/.conda/envs/pointpillars/lib/python3.7/site-packages/nuscenes/eval/detection/data_classes.py", line 155, in deserialize return cls(sample_token=content['sample_token'], TypeError: string indices must be integers

    for sample_token, boxes in content.items():
        eb.add_boxes(sample_token, [EvalBox.deserialize(box) for box in boxes])
    return eb
pyaf commented 5 years ago

use nuscenes v1.0.1