Soft-NMSでエラー

[kminoda]

https://github.com/Toma0916/GlobalWheatDetection/blob/master/experiment/manhattan/exp_02/config.json これとかででる

Traceback (most recent call last):
  File "train.py", line 249, in <module>
    trained_epoch=0)
  File "train.py", line 95, in exec_train
    evaluate_epoch(model, valid_data_loader, logger, optimizer)
  File "train.py", line 136, in evaluate_epoch
    processed_outputs = postprocessing(copy.deepcopy(preds), config["valid"]) if 'valid' in config.keys() else outputs
  File "/home/minoda/git/GlobalWheatDetection/utils/postprocess.py", line 317, in postprocessing
    outputs = ensemble_boxes_method_list[ensemble_boxes_method_name](copy.deepcopy(outputs), **config['post_processor']['config'])
  File "/home/minoda/git/GlobalWheatDetection/utils/postprocess.py", line 176, in soft_non_maximum_supression
    picked_boxes, picked_scores = non_maximum_supression_each(output['boxes'], output['scores'], threshold=None, sigma=sigma, method_type='soft')
  File "/home/minoda/git/GlobalWheatDetection/utils/postprocess.py", line 152, in non_maximum_supression_each
    confidence_score[np.where(index<order)[0]] = weights * confidence_score[np.where(index<order)[0]]
ValueError: operands could not be broadcast together with shapes (3,) (2,)

[Toma0916]

なおすわ

[kminoda]

あざぽよ