I'm looking to use an example of ml5, in a raspberry pi zero w. Is that possible?
I would need it for the Image Classification example for training and running the model.
I have tried running it on a Raspberry pi 3B and a Raspberry pi zero w with Raspbian on Chrome, but I was not able to load the example page. I'm guessing this might be because of the poor performance of the hardware.
Do you have any idea if I might need to do something differently or if the hardware is not powerful enough to run ml5?
I also meet the same issue. It's work with P5js show image in webcam. Model leaded but percent of confident is not update information.
Do you have check configuration to have run ml5js on raspberrry pi?
Thanks you
I think that this is probably a hardware problem rather than a problem with ml5 iteslf so this issue could be closed.
Thanks @EmmaGoodliffe for the feedback here!
There have been a lot of developments regarding hardware and ML. Hardware isn't my personal fort茅 but @yining1023 @oveddan @ellennickles have done a number of projects w/ ML and various hardware boards.
You can see some of Ellen's tutorial materials here:
https://github.com/ml5js/Intro-ML-Arts-IMA/tree/source/03_ml_models#session-b-physical-interaction
Interesting! I wonder if something like the Coral USB Accelerator would be helpful for this. I've at least had good experiences running TF Lite models on Coral Dev Boards.
Closing this issue for now. @dianebecheras please let us know if there's anything else we can do to help! Thanks again @EmmaGoodliffe for your input on this!