Person Detection Model

So, I got my hands on an OpenMV Cam H7 Plus. I have updated the latest OpenMV IDE and my firmware is at 4.4.1 (latest). I have been looking at the examples and I watched this video with interest:

I saw in a different post that the person_detection model is not included in the examples. So, I tried building a sketch which would load the person_detection model with this line:

# Load built-in FOMO face detection model
labels, net = tf.load_builtin_model("person_detection")

Unfortunately I am getting an error that the file is not found. I believe this is a built in model as indicated here:

So, can you please let me know what I am doing wrong? Do I need to load the model onto the microSD card and include a path? Is this model built in as you state in the YouTube video. I am excited about getting started so please excuse all the questions.



Hi, the video is out of date. We were shipping the Google person detector model on the camera back then. We just have the FOMO face model on the camera now. If you want the person detection model you can put that on your camera via a TensorFlow Lite file or you can build it into your firmware. I previously helped another user with this on the forums. You should be able to find my post about that.

@kwagyeman ,

Thank you - sorry if I did not make my question clear.

I was trying to follow the thread you had on this topic:

From what I saw, this model simply needed to be loaded using the command I referenced above. I have followed the links you provided and I don’t see why the person detection model is not being found.

I will, of course, keep looking. any clues appreciated.



Hi, you have to put the actual model file and labels in that folder and re-build the firmware. We don’t ship the person model with the firmware anymore.

1 Like