Faster rcnn 训练错误 InvalidArgumentError (see above for traceback): assertion failed: [] [Condition x =

 在尝试跑通tf-Faster-RCNN的时候，遇到了这个问题。

Traceback (most recent call last):
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1334, in _do_call
    return fn(*args)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1319, in _run_fn
    options, feed_dict, fetch_list, target_list, run_metadata)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1407, in _call_tf_sessionrun
    run_metadata)
tensorflow.python.framework.errors_impl.InvalidArgumentError: assertion failed: [] [Condition x == y did not hold element-wise:] [x (losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/Shape_1:0) = ] [101 1 1] [y (losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/strided_slice:0) = ] [101]
     [[{{node losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/assert_equal/Assert/Assert}}]]
     [[{{node losses/mul}}]]

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "faster_rcnn_conv5.py", line 297, in <module>
    main()
  File "faster_rcnn_conv5.py", line 290, in main
    model.train()
  File "faster_rcnn_conv5.py", line 170, in train
    summary = self._run_train_iter(feed_dict)
  File "faster_rcnn_conv5.py", line 138, in _run_train_iter
    summary, _ = self.sess.run([self.merged, self.optimizer], feed_dict=feed_dict)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 929, in run
    run_metadata_ptr)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1152, in _run
    feed_dict_tensor, options, run_metadata)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1328, in _do_run
    run_metadata)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/client/session.py", line 1348, in _do_call
    raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.InvalidArgumentError: assertion failed: [] [Condition x == y did not hold element-wise:] [x (losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/Shape_1:0) = ] [101 1 1] [y (losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/strided_slice:0) = ] [101]
     [[node losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/assert_equal/Assert/Assert (defined at ../Lib/loss_functions.py:91) ]]
     [[node losses/mul (defined at faster_rcnn_conv5.py:114) ]]

Caused by op 'losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/assert_equal/Assert/Assert', defined at:
  File "faster_rcnn_conv5.py", line 297, in <module>
    main()
  File "faster_rcnn_conv5.py", line 288, in main
    model = FasterRcnnConv5(flags, dictionary)
  File "faster_rcnn_conv5.py", line 40, in __init__
    super().__init__(flags_input, flags_input['run_num'], vram=cfg.VRAM, restore=flags_input['restore_num'])
  File "../Lib/TensorBase/tensorbase/base.py", line 677, in __init__
    self._optimizer()
  File "faster_rcnn_conv5.py", line 113, in _optimizer
    self.fast_rcnn_cls_loss = self.fast_rcnn_net['TRAIN'].get_fast_rcnn_cls_loss()
  File "../Networks/faster_rcnn_networks.py", line 245, in get_fast_rcnn_cls_loss
    return fast_rcnn_cls_loss(fast_rcnn_cls_score, labels)
  File "../Lib/loss_functions.py", line 91, in fast_rcnn_cls_loss
    fast_rcnn_cross_entropy = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits=tf.squeeze(fast_rcnn_cls_score), labels=labels))
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/ops/nn_ops.py", line 2670, in sparse_softmax_cross_entropy_with_logits
    array_ops.shape(logits)[:-1]))
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/ops/check_ops.py", line 512, in assert_equal
    return control_flow_ops.Assert(condition, data, summarize=summarize)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/util/tf_should_use.py", line 193, in wrapped
    return _add_should_use_warning(fn(*args, **kwargs))
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/ops/control_flow_ops.py", line 160, in Assert
    return gen_logging_ops._assert(condition, data, summarize, name="Assert")
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/ops/gen_logging_ops.py", line 72, in _assert
    name=name)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/op_def_library.py", line 788, in _apply_op_helper
    op_def=op_def)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/util/deprecation.py", line 507, in new_func
    return func(*args, **kwargs)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/ops.py", line 3300, in create_op
    op_def=op_def)
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/ops.py", line 1801, in __init__
    self._traceback = tf_stack.extract_stack()

InvalidArgumentError (see above for traceback): assertion failed: [] [Condition x == y did not hold element-wise:] [x (losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/Shape_1:0) = ] [101 1 1] [y (losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/strided_slice:0) = ] [101]
     [[node losses/fast_rcnn_cls_loss/SparseSoftmaxCrossEntropyWithLogits/assert_equal/Assert/Assert (defined at ../Lib/loss_functions.py:91) ]]
     [[node losses/mul (defined at faster_rcnn_conv5.py:114) ]]

一开始闭着眼就开始上网搜索，结果啥也没找到，然后看了一下问题原因，发现还挺简单的。

错误的地方在于，计算loss的时候数组维度对不上，导致他计算不了，到对应的位置改一下数组维度就完事了。

修改位置：Lib/loss_function.py

修改函数：fast_rcnn_cls_loss，大约在80行。

修改内容：将函数的

fast_rcnn_cross_entropy = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits=tf.squeeze(fast_rcnn_cls_score), labels=labels))

修改为：

fast_rcnn_cross_entropy = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits=tf.expand_dims(tf.squeeze(fast_rcnn_cls_score),1), labels=labels))

好了，其实就是增加了一维。

以后要先分析错误，不能闭着眼就想剽窃别人的成果，这样不好。