CIFAR10-ObjectDetection with ResNet50 achieves 83.2% accuracy on the test set & 89.5% on the train set
CIFAR-10 contains images of 10 classes which need to be classified
The images are low resolution 32*32 rgb images making the task even difficult
The .ipynb file contains implementation of ResNet50 inspired architecture
It achieves 83.2% accuracy on the test set after 20 epochs but is subject to a bit of variation each time due to the randomness involved in Stochastic Gradient descent