tensorflow: Error converting the model to TF Lite

System information

• Have I written custom code (as opposed to using a stock example script provided in TensorFlow): yes
• OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntu 16.04.4
• TensorFlow installed from (source or binary): source
• TensorFlow version (use command below): r1.6 commit: cbc658095ae228f2f557af47e4901d552573aa15
• Python version: 3.5.2
• Bazel version (if compiling from source): 0.11.1
• GCC/Compiler version (if compiling from source): gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.9)
• CUDA/cuDNN version: N/A (build without support CUDA)
• GPU model and memory: N/A (build without support CUDA)
• Exact command to reproduce:

Describe the problem

Trained model, successfully froze, it works on the tensorflow android, using TensorFlowInferenceInterface. I try to convert this into a TF Lite format, but I get an error.

Source code / logs

bazel-bin/tensorflow/contrib/lite/toco/toco \
    --input_file=./test_model/frozen_graph.pb \
    --input_format=TENSORFLOW_GRAPHDEF \
    --output_file=./test_model/unet.tflite \
    --output_format=TFLITE \
    --input_array='input' \
    --input_data_type=FLOAT \
    --input_shape=2,192,320,1 \
    --inference_type=FLOAT \
    --inference_input_type=FLOAT \
    --output_array='final/Sigmoid' \
    --v=1

2018-03-13 21:07:12.711948: I tensorflow/contrib/lite/toco/graph_transformations/graph_transformations.cc:39] Before Removing unused ops: 282 operators, 479 arrays (0 quantized)
2018-03-13 21:07:12.716274: I tensorflow/contrib/lite/toco/graph_transformations/graph_transformations.cc:39] Before general graph transformations: 282 operators, 479 arrays (0 quantized)
2018-03-13 21:07:12.716893: F tensorflow/contrib/lite/toco/graph_transformations/resolve_batch_normalization.cc:86] Check failed: mean_shape.dims() == multiplier_shape.dims()
Aborted (core dumped)