Not all classification models can detect imagenet dataset, for example resnet works great on ImageNet but resnet-caffe doesn't work at all

Ask a Question

Question

Not all classification models can detect imagenet dataset, for example resnet model works great on ImageNet but resnet-caffe model doesn't work at all

Further information

Relevant Area (e.g. model usage, backend, best practices, pre-/post- processing, converters):
Test image:ILSVRC2012_val_00007667.JPEG
Top5 for resnet50-v1-7

Top5 for resnet50-caffe2-v1-9

Is this issue related to a specific model?
Model name (e.g. mnist):
"bvlcalexnet-9",
"caffenet-9",
"densenet-9",
"efficientnet-lite4-11",
"googlenet-9",
"inception-v1-9",
"inception-v2-9",
"mobilenetv2-7",
"rcnn-ilsvrc13-9",
"resnet18-v1-7",
"resnet18-v2-7",
"resnet34-v1-7",
"resnet34-v2-7",
"resnet50-v1-7",
"resnet50-v2-7",
"resnet101-v1-7",
"resnet101-v2-7",
"resnet152-v1-7",
"resnet152-v2-7",
"resnet50-caffe2-v1-9",
"shufflenet-9",
"shufflenet-v2-10",
"squeezenet1.0-9",
"squeezenet1.1-7",
"vgg16-7",
"vgg16-bn-7",
"vgg19-7",
"vgg19-bn-7",
"vgg19-caffe2-9",
"zfnet512-9"
Model opset (e.g. 7):>7

Notes

Any additional information, code snippets.
This collection of models take images as input, then classifies the major objects in the images into 1000 object categories such as keyboard, mouse, pencil, and many animals.

The issue names no files or tests; start by reproducing the result with ILSVRC2012_val_00007667.JPEG across the listed classification models, especially resnet50-v1-7 and resnet50-caffe2-v1-9. Done means identifying why the model outputs differ or fail and documenting the confirmed behavior or required correction.