ABOUT THE PROJECT
LittlePlaNet is a Convolutional Neural Network built to recognize the location of Google StreetView images taken in ten different cities. The neural network achieved over 70% test accuracy on the dataset, more that three times better than human-level performance. We also used and explored visualization techniques that allowed us to understand the inner workings of the neural network.
Developed For
This project was developed for CS231n at Stanford University.