In my last post I described how to use Python’s numpy library to classify an image as being pizza or not pizza. That blog post can be found here:- https://medium.com/@tracyrenee61/use-numpy-to-classify-an-image-as-being-pizza-or-not-pizza-f25c7c3fe9b7
In this blog post I have translated the code in numpy to Jax, which is a python library used for research purposes. Although Jax code is similar to numpy in many ways, it is not a direct translation. For instance, Jax arrays are immutable so they are more complex to work with. Jax’s random numbers are also different from numpy’s, so this requires a change in the coding.
The dataset that I have used is Kaggle’s pizza or not pizza dataset and it can be found here:- https://www.kaggle.com/datasets/carlosrunner/pizza-not-pizza/data
I have written the classification program in Python using Kaggle’s free online Jupyter Notebook.
Once I created the Jupyter Notebook, I imported the libraries that I would need to execute the program, being:-
- Jax to create the neural network and perform numerical computations,
- PIL to process the images and view them,
- Pathlib to establish paths of the images,
- Cv2 to carry out computer vision…
