Hi everyone,
I successfully followed the tutorial to create a custom object detection model using Edge Impulse, and it worked well on my mobile phone.
As a proof of concept, I’m now trying to deploy and run the same model on a Windows laptop using the C++ library, following the documentation here: As a generic C++ library | Edge Impulse Documentation.
In my model dashboard, I used the deployment option to download the C++ library of the model files. The library includes several folders:
- edge-impulse-sdk/ - The main toolkit with all the AI functionality
- model-parameters/ - Settings for your specific model
- tflite-model/ - Your actual trained AI model
However, I’m running into issues when trying to build the sample code from this GitHub repo:
GITHUB /edgeimpulse/example-standalone-inferencing
Using this include line:
#include “edge-impulse-sdk/classifier/ei_run_classifier.h”
Visual Studio throws hundreds of build errors, even though I’ve configured the C/C++ include directories and other project settings as described in the documentation.
Has anyone successfully built the C++ library from a test or custom model using Visual Studio or VS Code?
Any tips or working configurations would be greatly appreciated!
Thanks in advance,
Felix
Public project at: Simple Obj Detection with Edge Impulse and a mobile phone - Deployment - Edge Impulse
C/C++ → General → Additional Include Directories: Add paths to the folders containing header files.
$(ProjectDir)$(ProjectDir)model-parameters/
$(ProjectDir)edge-impulse-sdk/
$(ProjectDir)tflite-model/
C/C++ → Preprocessor → Preprocessor Definitions:
EIDSP_QUANTIZE_FILTERBANK=0
EI_CLASSIFIER_TFLITE_ENABLE_CMSIS_NN=0
EI_CLASSIFIER_TFLITE_ENABLE_ESP_NN=0
EI_CLASSIFIER_TFLITE_ENABLE_ARC=0
TF_LITE_DISABLE_X86_NEON=1
EI_CLASSIFIER_ALLOCATION_STATIC
EI_CLASSIFIER_ENABLE_DETECTION_POSTPROCESS_OP=1
NDEBUG