The objective of this project was to build a model to identify and classify images of numbers potrayed by human hands.The same model was then ported to the MNIST dataset for digit recognition using neural nets.
1.Built a neural net using tensorflow and keras along with other python libraries to classify hand made images of numbers from one to six using adam optimizer along with softmax regression.
2.CNN model to make a classifier for the same problem in tensorflow and keras.
3.A residual net with 50 layers.
All this was done in order to observe the increase in accuracy with the addition of more accurate and sophisticated ML methods.