Artificial Neural Networks Project: Street View Housing Number Digit Recognition

Classification using Artificial Neural Networks using the treet View Housing Numbers (SVHN) image dataset.

This uses deep learning to recognize objects in natural scenes.

The SVHN dataset contains over 600,000 labeled digits cropped from street-level photos. It has been used in neural networks created by Google to improve the map quality by automatically transcribing the address numbers from a patch of pixels. The transcribed number with a known street address helps pinpoint the location of the building it represents.

This is a feed-forward neural network model that can recognize the digits in the images.