2019-08-15 08:47 AM
Hello everyone,
It's good to see a number of tutorials/implementations posted by community members which encourage beginners like me to implement AI at the microcontroller level.
Is there any complete image classification example or tutorial where ST microcontroller capture images via DCMI interface and immediately classify images using the pre-trained network?
If available, I sincerely request to share the link.
Thanks in advance,
M
2019-08-16 01:16 AM
Hello,
The new function pack FP-AI-VISION1 is providing what you are looking for: a Computer Vision AI example of food classification from DCMI camera.
The function pack is based on STM32H747 Discovery board and you can find it here:
https://www.st.com/content/st_com/en/products/embedded-software/mcu-mpu-embedded-software/stm32-embedded-software/stm32cube-expansion-packages/fp-ai-vision1.html
Best Regards,
Matthieu
2019-08-16 07:48 AM
Hello Matthieu,
Thank you for your reply. yes, this is a great starting point.
Is the Keras CNN model employed in these application open source? I mean, I'll use custom images to re-train the model.
May be your model (github project) would be helpful to follow the steps and implement a new model for the custom image set.
Thank you.
2019-08-17 07:28 AM
Hello,
A Keras CNN model is provided (under ST free license) as part of the package as well as the quantization scripts so you can retrain and quantize on your own image dataset?
Best Regards
Matthieu