diff options
| author | Narumol Prangnawarat <narumol.prangnawarat@arm.com> | 2019-02-25 17:26:05 +0000 |
|---|---|---|
| committer | Narumol Prangnawarat <narumol.prangnawarat@arm.com> | 2019-02-26 17:41:15 +0000 |
| commit | 4628d05455dfc179f0437913185e76888115a98a (patch) | |
| tree | a8eac68ee5aee88a7071ac6f13af7932b98caa87 /src/armnnTfLiteParser/TfLiteParser.hpp | |
| parent | 452869973b9a45c9c44820d16f92f7dfc96e9aef (diff) | |
| download | armnn-4628d05455dfc179f0437913185e76888115a98a.tar.gz | |
IVGCVSW-2560 Verify Inference test for TensorFlow Lite MobileNet SSD
* Assign output shape of MobileNet SSD to ArmNN network
* Add m_OverridenOutputShapes to TfLiteParser to set shape in GetNetworkOutputBindingInfo
* Use input quantization instead of output quantization params
* Correct data and datatype in Inference test
Change-Id: I01ac2e07ed08e8928ba0df33a4847399e1dd8394
Signed-off-by: Narumol Prangnawarat <narumol.prangnawarat@arm.com>
Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Diffstat (limited to 'src/armnnTfLiteParser/TfLiteParser.hpp')
| -rw-r--r-- | src/armnnTfLiteParser/TfLiteParser.hpp | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/src/armnnTfLiteParser/TfLiteParser.hpp b/src/armnnTfLiteParser/TfLiteParser.hpp index 3fe4809aa2..2895487214 100644 --- a/src/armnnTfLiteParser/TfLiteParser.hpp +++ b/src/armnnTfLiteParser/TfLiteParser.hpp @@ -193,6 +193,10 @@ private: /// Connections for tensors in each subgraph /// The first index is the subgraph ID, the second index is the tensor ID std::vector<TensorConnections> m_SubgraphConnections; + + /// This is used in case that the model does not speciry the output. + /// The shape can be calculated from the options. + std::vector<std::vector<unsigned int>> m_OverridenOutputShapes; }; } |