The image processing of inception_v2

[JerryShih]

I try to use caffe2 onnx backend to run the opset8 inception_v2 model. The caffe2 onnx backend running with sample test data is 99% similar with the expected test result, but the accuracy rate is 0 when I run with ILSVRC2012 dataset. There is no image processing step in this link. Maybe it should have some pre-processing steps before running inference?

@houseroad @bddppq

[santoshchilkunda]

Facing the same issue. While sample test cases match the reference data, the accuracy is 0. By looking the input_*.pb, the data range seems to be 0-255 float. I set the same (i.e. no preprocessing).

Has anybody been able to get the quoted accuracy for inception_v2?

[gcunhase]

Any update on this?