Question

我按照https://github.com/deboc/py-faster-rcnn/blob/master/help/Readme.md的指示训练py-faster-rcnn 在自定义数据集上。

但是，我收到以下错误：

    Preparing training data...
    Process Process-1:
    Traceback (most recent call last):
  File "/usr/lib/python2.7/multiprocessing/process.py", line 258, in _bootstrap
    self.run()
  File "/usr/lib/python2.7/multiprocessing/process.py", line 114, in run
    self._target(*self._args, **self._kwargs)
  File "./train_faster_rcnn_alt_opt.py", line 122, in train_rpn
    roidb, imdb = get_roidb(imdb_name)
  File "./train_faster_rcnn_alt_opt.py", line 67, in get_roidb
    roidb = get_training_roidb(imdb)
  File "/home/Work/code/py-faster-rcnn/tools/../lib/fast_rcnn/train.py", line 122, in get_training_roidb
    rdl_roidb.prepare_roidb(imdb)
  File "/home/Work/code/py-faster-rcnn/tools/../lib/roi_data_layer/roidb.py", line 31, in prepare_roidb
    gt_overlaps = roidb[i]['gt_overlaps'].toarray()
AttributeError: 'numpy.ndarray' object has no attribute 'toarray'

这是roidb.py的代码片段（第31行）：

for i in xrange(len(imdb.image_index)):
    roidb[i]['image'] = imdb.image_path_at(i)
    roidb[i]['width'] = sizes[i][0]
    roidb[i]['height'] = sizes[i][1]
    # need gt_overlaps as a dense array for argmax
    gt_overlaps = roidb[i]['gt_overlaps'].toarray()

我无法找到解决方法。

Answer 1

在标准实现中，发布行上的roidb[i]['gt_overlaps']应该是稀疏矩阵。您可能在定义自己的数据集地面实况读取器时忘记了这一点，因为它最初是一个numpy.ndarray，后来被转换。其他解决方法是可能的（并且可能更好，具体取决于应用程序）但如果您希望保持尽可能接近py-faster-rcnn，只需将每个基础事实的gt_overlaps键包裹起来，如overlaps = scipy.sparse.csr_matrix(overlaps)，引用可以在原始数据集文件中找到（即lib/datasets/pascal_voc.py）。

AttributeError：'numpy.ndarray'对象在py-faster-rcnn中没有属性'toarray'

1 个答案: