Tensorflow lite model used to classify images in real time of violin playing using the openMV h7 plus camera/microcontroller
Example of images along with their outputs (as color frames) is in the example_video_320x280.mp4 video. Some more examples are in the video_examples folder.
To use: Download 200x200 data. Consider adding your own 200x200 data of images of violin playing, using the python scrips from code -> frames_into_folder.py.
Then, Download image data onto google drive, as google colab is quite useful here.
From the code folder, open either most_recent.ipynb or old_tftest3classes.ipynb. Then simply run this code to create the tflite model.
Run main.py with the tflite model on the openMV camera.
With any questions, email me @ enla2020@mymail.pomona.edu
Thanks to Suren Jayasuriya, Odrika Iqbal, Seth Thorn, Byron Lahey as a part of the Visual Media REU at ASU Special thanks to @CzJLee for help cleaning up code