Run a simple Core ML classification model without using coremlc generated wrapper
- MobileNet_EdgeTPU_multi_arrays.mlmodel: converted from MobileNet EdgeTPU model, Both input and output types are MLMultiArray
- grace_hopper.raw: 24-bit rgb raw image, converted with something like
convert grace_hopper.bmp -gravity Center -resize 224x224 -extent 224x224 foo.rgb
make run
You should get something like
2022-07-10 21:16:06.934 test_cmd[91957:28263153] 653
653 means the (653+1)-th class, which is "military uniform". Since this image is Rear Admiral Grace Hopper in military uniform, this is an expected result.