inference-engine/tests/unit/builders/batch_normalization_layer_test.cpp

   1 // Copyright (C) 2018-2019 Intel Corporation
   2 // SPDX-License-Identifier: Apache-2.0
   3 //
   4
   5 #include <gtest/gtest.h>
   6 #include <string.h>
   7 #include <ie_builders.hpp>
   8 #include <builders/ie_batch_normalization_layer.hpp>
   9
  10 #include "builder_test.hpp"
  11
  12 using namespace testing;
  13 using namespace InferenceEngine;
  14
  15 class BatchNormalizationLayerBuilderTest : public BuilderTestCommon {};
  16
  17 //TEST_F(BatchNormalizationLayerBuilderTest, cannotCreateBatchNormalizationWithoutWeightOrBiases) {
  18 //    ASSERT_THROW(((Builder::Layer)Builder::BatchNormalizationLayer("in1")), InferenceEngine::details::InferenceEngineException);
  19 //    ASSERT_THROW(((Builder::Layer)Builder::BatchNormalizationLayer("in1")
  20 //            .setWeights(generateBlob(Precision::FP32, {3}, Layout::C))), InferenceEngine::details::InferenceEngineException);
  21 //    ASSERT_THROW(((Builder::Layer)Builder::BatchNormalizationLayer("in1")
  22 //            .setBiases(generateBlob(Precision::FP32, {3}, Layout::C))), InferenceEngine::details::InferenceEngineException);
  23 //}
  24
  25 TEST_F(BatchNormalizationLayerBuilderTest, getExistsLayerFromNetworkBuilder) {
  26     Builder::Network network("Test");
  27     idx_t weightsId = network.addLayer(Builder::ConstLayer("weights").setData(generateBlob(Precision::FP32, {3}, Layout::C)));
  28     idx_t biasesId = network.addLayer(Builder::ConstLayer("biases").setData(generateBlob(Precision::FP32, {3}, Layout::C)));
  29     Builder::BatchNormalizationLayer bnBuilder("bn");
  30     idx_t bnId = network.addLayer({{0}, {weightsId}, {biasesId}}, bnBuilder);
  31     Builder::BatchNormalizationLayer bnBuilderFromNetwork(network.getLayer(bnId));
  32     ASSERT_EQ(bnBuilderFromNetwork.getEpsilon(), bnBuilder.getEpsilon());
  33     bnBuilderFromNetwork.setEpsilon(2);
  34     ASSERT_NE(bnBuilderFromNetwork.getEpsilon(), bnBuilder.getEpsilon());
  35     ASSERT_EQ(bnBuilderFromNetwork.getEpsilon(), network.getLayer(bnId)->getParameters()["epsilon"].as<float>());
  36 }