WATai-open-challenge-2023

There are 2 image classification challenges to complete using the Stanford Online Products dataset. The input images are 64x64 pixels with 3 channels (RGB), each taking an integer value between 0-255.

Challenge 1: using classical ML (and scikit-learn) to distinguish between bikes and cabinets

I experimented with and learned about many of the algorithms available on scikit-learn. Tuned a model to an accuracy of ~90%.

Challenge 2: using neural networks to sort between 12 different items

Implemented a ResnetV2 model to reach an accuracy of ~63%.