too many indices for array

[jiamingNo1]

arm_source_layers = ['conv3_3', 'conv4_3', 'conv5_3', 'fc7', 'conv6_2'] odm_source_layers = ['P2', 'P3', 'P4', 'P5', 'P6']

多了一个特征层，然后训练的时候没有报错，测试的时候出现

Traceback (most recent call last): File "test/refinedet_test.py", line 39, in single_scale_test_net(net, imdb, targe_size=input_size) File "/opt/caffe/test/lib/fast_rcnn/test.py", line 247, in single_scale_test_net imdb.evaluate_detections(all_boxes, output_dir) File "/opt/caffe/test/lib/datasets/pascal_voc.py", line 302, in evaluate_detections self._do_python_eval(output_dir) File "/opt/caffe/test/lib/datasets/pascal_voc.py", line 271, in _do_python_eval use_07_metric=use_07_metric) File "/opt/caffe/test/lib/datasets/voc_eval.py", line 148, in voc_eval BB = BB[sorted_ind, :] IndexError: too many indices for array 请问该如何解决？

[sfzhang15]

@ouchjm 只根据目前这些信息我也不好判断是哪里的问题，你可以参考pdb调试进行单步调试看看问题在哪里。

[jiamingNo1]

@sfzhang15 谢谢学长，我已经按照SSD的测试方法解决了。