Help with Nano 33 BLE Computer Vision Code

[jhertz2]

Hi all,

I'm trying to modify the person_detection example from Harvard TinyMLx for the arduino nano ble 33 to use my own model instead of the one provided.

I used Teachable Machine to create a TensorFlow Lite Model, converted that to a c file using
xxd -i converted_model.tflite > model_data.cc

Then I took that model_data.cc, inserted it in place of the the model from the harvard code (and changed names accordingly).

Now, when I try to upload my sketch, I get the following error:

c:/users/jacob.hertz/appdata/local/arduino15/packages/arduino/tools/arm-none-eabi-gcc/7-2017q4/bin/../lib/gcc/arm-none-eabi/7.2.1/../../../../arm-none-eabi/bin/ld.exe: region `FLASH' overflowed by 15208 bytes

collect2.exe: error: ld returned 1 exit status

Using library Arduino_TensorFlowLite at version 2.4.0-ALPHA in folder: C:\Users\jacob.hertz\Documents\Arduino\libraries\Arduino_TensorFlowLite

Using library Harvard_TinyMLx at version 1.0.1-Alpha in folder: C:\Users\jacob.hertz\Documents\Arduino\libraries\Harvard_TinyMLx

Using library Wire in folder: C:\Users\jacob.hertz\AppData\Local\Arduino15\packages\arduino\hardware\mbed_nano\2.0.0\libraries\Wire (legacy)

exit status 1

Error compiling for board Arduino Nano 33 BLE.

Can anyone help out?

[sterretje]

@jhertz2, your topic was moved to a more suitable location on the forum.

[Klaus_K]

Welcome to the forum.

@jhertz2

`FLASH’ overflowed by 15208 bytes

Does sound like you do not have enough Flash memory. Any chance your model is too large?

I have not work with this before, any getting started you can recommend?