MAIS Hack 2020
We often go outside and get confused by the animals we see. So we decided to use neural networks and machine learning to fix this complex problem.
We learned how neural networks and especially convolutional neural networks worked, how to implement them using pytorch, and how to manipulate a dataset to get it into the shape we needed for training.
We also learned that some problems are a lot more complex to solve than others, as we initially tried to get our model to recognize different cat breeds, but only got a 10% accuracy rate. In the end, we got a 54% accuracy on ten different animal classes.