Export or edit this event...

Using Deep Learning techniques and Tensorflow for Image Classification



Syed Nasar, a recent Masters in ML grad from Georgia Tech, will walk us through an Image Classification demo using CNN and Tensorflow. Currently, Syed is doing two Nanodegrees on Deep Learning and Artificial Intelligence. He has worked with some of the cutting edge technologies and leading researchers at Georgia Tech, Google, NVidia, and Udacity.

In this prototype, Syed will classify images from the CIFAR-10 dataset. The dataset consists of airplanes, dogs, cats, and other objects. Syed plans to preprocess the images, then train a convolutional neural network on all the samples. In this process Syed will explain how to build a convolutional NN. Explain concepts around image data preparation (one-hot encoding etc.). Explain concepts like max pooling, dropout, and fully connected layers. At the end, Syed will showcase how the neural network is able to predict on the sample images.