This repo contains a simple Resnet18 classifier (trained on a Macbook without a GPU) to differentiate Horses from Donkeys (because why not?). There are two different models :
- A very normal, everyday binary classification model
- Same as
1
, but a singe channel in every image is randomly ignored while training and testing the model.
The data for this project was scraped from Google Images