diff options
author | Matthew Bentham <matthew.bentham@arm.com> | 2018-10-29 17:39:49 +0000 |
---|---|---|
committer | Matthew Bentham <matthew.bentham@arm.com> | 2018-10-29 17:41:53 +0000 |
commit | 4322d36a2d6d9fca16a661019b8c5dac0c1e81ec (patch) | |
tree | 72fdebd79624f268707cc640d1f153b5d2341bd1 /tests/TfLiteVGG16Quantized-Armnn/TfLiteVGG16Quantized-Armnn.cpp | |
parent | 382a91d5029e83002bda4ab006f9c73340d679fe (diff) | |
download | armnn-4322d36a2d6d9fca16a661019b8c5dac0c1e81ec.tar.gz |
IVGCVSW-2029 Tweak results handling for batch size 2 test
When looking for the top probability, use the 'first' result not the 'second'.
This avoids an issue where for batched tests the classification index was reported wrongly.
Still doesn't correctly handle multiple results with the exact probabibility, or batched testing,
but it's slightly more correct than before.
Change-Id: I57d33552754667613e222d9d2037e12c87a96854
Diffstat (limited to 'tests/TfLiteVGG16Quantized-Armnn/TfLiteVGG16Quantized-Armnn.cpp')
-rw-r--r-- | tests/TfLiteVGG16Quantized-Armnn/TfLiteVGG16Quantized-Armnn.cpp | 1 |
1 files changed, 0 insertions, 1 deletions
diff --git a/tests/TfLiteVGG16Quantized-Armnn/TfLiteVGG16Quantized-Armnn.cpp b/tests/TfLiteVGG16Quantized-Armnn/TfLiteVGG16Quantized-Armnn.cpp index 1313d2d01a..acaabe4487 100644 --- a/tests/TfLiteVGG16Quantized-Armnn/TfLiteVGG16Quantized-Armnn.cpp +++ b/tests/TfLiteVGG16Quantized-Armnn/TfLiteVGG16Quantized-Armnn.cpp @@ -16,7 +16,6 @@ int main(int argc, char* argv[]) // Coverity fix: The following code may throw an exception of type std::length_error. std::vector<ImageSet> imageSet = { - // Class number in probability print out offset by 1000 due to batch size fix {"Dog.jpg", 669}, {"Cat.jpg", 669}, {"shark.jpg", 669}, |