Train error.

[herrickli]

Hi, I use instruction train_cascade_rcnn.ipynb to trian cascade rcnn on my custom dataset, and I got this error

Traceback (most recent call last):
  File "train.py", line 33, in <module>
    _ = model((batch_imgs, batch_metas), training=False)
  File "/cascade-rcnn-tf2.2/detection/models/detectors/cascade_rcnn.py", line 208, in __call__
    proposals = transforms.delta2bbox(rois, rcnn_deltas, self.RCNN_TARGET_MEANS, self.RCNN_TARGET_STDS)
  File ''/cascade-rcnn-tf2.2/detection/core/bbox/transforms.py", line 72, in delta2bbox
    center_y += delta[:, 0] * height
  File "/usr/local/lib/python3.7/site-packages/tensorflow_core/python/ops/math_ops.py", line 899, in binary_op_wrapper
    return func(x, y, name=name)
  File "/usr/local/lib/python3.7/site-packages/tensorflow_core/python/ops/math_ops.py", line 1206, in _mul_dispatch
    return gen_math_ops.mul(x, y, name=name)
  File "/usr/local/lib/python3.7/site-packages/tensorflow_core/python/ops/gen_math_ops.py", line 6698, in mul
    _six.raise_from(_core._status_to_exception(e.code, message), None)
  File "<string>", line 3, in raise_from
tensorflow.python.framework.errors_impl.InvalidArgumentError: Incompatible shapes: [2000,4] vs. [2000] [Op:Mul] name: mul/

There are 8 classes in my custom dataset, the delta in delta2bbox function has shape of (2000,9,4), how should I fix this error to train my own dataset, thanks a lot.

[HirataYurina]

Maybe you need to set a background label in you label file.

[HirataYurina]

Have fixes this bugs and update your codes, bro.