Automatic Image Filtering Using Deep Learning
Two Hat Security is a company that develops next generation moderation tools for social networking apps. Since images are of the most important data shared by social networking apps, an important problem for the company is to identify images that are unsafe or inappropriate. In particular, images containing certain objects (e.g. knife, gun, bikini, etc.) are considered unsafe. It is obviously not practical to manually sift through all the images to find the unsafe ones. The objective of this project is to develop algorithms to automatically filter images and identify those that are considered unsafe. We will develop algorithms using deep learning (in particular, convolutional neural network) to identify whether an image contains any of the objects that are considered to be unsafe.