This is a deep learing approach to Udacitys "Vehicle Detection and Tracking" project in the CarND using CNNs. Just follow the ipython notebook. The the the dataset this was model was trained on can be found in udacity's repo: https://github.com/udacity/CarND-Vehicle-Detection
Specifically I used: vehicles non-vehicles and some additional samples generated from the GTI vehicle image database