AI-based image classifer which detetcs common road objects; automobiles, trucks, bicycles, buses, motorcycles, and pickup trucks.

The training module is run from vehicle_classifier.py

Source datasets are CIFAR10 and CIFAR100: https://www.cs.toronto.edu/~kriz/cifar.html?fbclid=IwAR1z93qFhvoxx4F1AxjWHcLQSV42785c5wUX FCiqAH6qTKK3lJFL_Ec-_Nc

CIFAR-10: 60,000 images overall, of which a total of 1,200 will be used for the system. 10 Classes overall, of which two will be used. These classes are “automobiles” and “trucks”. Each class contains 6,000 images; however, we will only use 600 each from both classes. Out of the 600, 500 will be used for training, and 100 will be used for testing. Each image is coloured and has dimensions 32x32 pixels.

CIFAR-100: 60,000 images overall, however these images come in sets of 600 per class. Total of 100 classes in the dataset, of which four classes will be used. Similarly, the 600 images will consist of 500 training images and 100 testing images. These images are also coloured and have dimensions of 32x32 pixels.

**Lookup-table for interpreting confusion matricies: (class:row/col) {car: 0, truck: 1, bicycle: 2, bus: 3, motorcycle: 4, pickup: 5}