I tried following this example but, the face detection model is not in my IDE under Machine Vision Examples as it is with yours.
How do I add the face detection application so I can test the Smile Detection TFL library I created using Edge Impulse?
Thanks, Chip
Hi, we have the FOMO face detection model in the firmware and the haar cascade. Both are by default in the firmware.
Thank you for your response. I am sure I am missing something but when I follow your example video, I download the OpenMV Library and take the labels.txt and trained.tflite files and drag them onto my OpenMV Cam H7 Plus’s microSD card.
Then, I go back to your video and I see you open the image classification model from the Examples / Machine Vision menu:
From this menu, you select “tf_face_detection”. I don’t have this option, here is what my machine learning menu looks like:
I can select tf_image_classification but it will not work with the files I provided (even if I rename them to match the expected file name in the python code).
I have the latest OpenMV IDE software and the latest firmware on my camera. I am working with my daughter on a summer project and we were hoping to get through this example to understand the process before moving onto the FOMO model application.
What am I missing?
Thanks,
Chip
Hi, we never shipped a smile detection model. It was just a demo that I made a video with to show off that you could train custom models using Edge Impulse.
You need to follow this: https://docs.edgeimpulse.com/docs/deployment/running-your-impulse-openmv
Please note that the videos we post just capture a snapshot in time of the state of the project.
Thank you for that. We were able to get the OpenMV Library file to work just fine. The OpenMV Firmware did not work but as long as we have a way forward, it is all good.
Point taken on the videos, glad you are making so much progress with this platform.
On to FOMO…
Thanks,
Chip
