I am facing the issue with the Grove Vision AI camera module. I do not know if there is a possibility to combine the Trained model with Arduino code.
Actually, I followed some tutorials on Seeed wiki which focuses on trained models and deploying that model to the device by coping the files of the trained model to the drive Grove AI
What I want is the code of Arduino or some standalone C++ code which runs on the Grove Vision module and sends some of the detection results via interface such as Grove etc.
So do you know is it yet possible with Grove Vision AI module or some third party software can help in this case??
In fact, we are in talks to support Edge Impulse and are pleased that Edge Impulse will complete support for Vision AI as soon as possible, so when that happens, the output of the results from the models will be done through Arduino programming, so stay tuned.
Do you have some C++ interface to detect for example wild fires and send the result via some grove interface? is it possible?
Hi, as this product is currently designed, to recognise other models you may need to train them yourself or download them from some model websites (of course, we will go on to do this piece specifically in the future). Vision AI can then output the number of models identified to your model and its confidence level. The output is all done in C/C++.
For more information, please refer to the Wiki below.
I have followed that tutorial and deployed the model to detect wild fires to the Grove AI module.
I have one question that weather the model installed once will remain active on next power-up?
I have seen some wrong results. Thats is why asking.
The model firmware should be retained by the power off.