emza_yaw_landmarks_fvp

To run the example, clone the ML repo from ARM:

mkdir ~/demo

cd ~/demo

git clone -b 22.02 https://review.mlplatform.org/ml/ethos-u/ml-embedded-evaluation-kit

cd ml-embedded-evaluation-kit

git checkout -b test_branch ed35a6fea4a1604db81c56fc71f7756822fcf212

clone this repo:

cd ~/demo

git clone https://github.com/emza-vs/emza_yaw_landmarks_fvp.git

Merge the ml-embedded-evaluation-kit folder from emza_yaw_landmarks_fvp into the ml-embedded-evaluation-kit folder from ARM,overwrite the files.

Now go to the modified ml-embedded-evaluation-kit folder and build the example

cd ~/demo/ml-embedded-evaluation-kit

./download_dependencies.py

mkdir build

cd build

cmake .. -DUSE_CASE_BUILD=object_detection -Dobject_detection_IMAGE_SIZE=160 -Dobject_detection_MODEL_TFLITE_PATH=resources_downloaded/object_detection/ssd_slim_120x160x1_yaw_landmarks_v3_int8_vela_H256.tflite -DTARGET_PLATFORM=mps3 -DTARGET_SUBSYSTEM=sse-300 -DCMAKE_TOOLCHAIN_FILE=scripts/cmake/toolchains/bare-metal-gcc.cmake

make

run the FVP

~/FVP_Corstone_SSE-300/models/Linux64_GCC-6.4/FVP_Corstone_SSE-300_Ethos-U55 -C ethosu.num_macs=256 -a ./bin/ethos-u-object_detection.axf

NOTE: for detailed step-by step instruction to set-up the GCC and Python toolchanin please look at the readme file here:

https://github.com/emza-vs/face_detection_example_arm_u55/blob/master/README.md