Emotion recognition in images: from idea to an AI model in production

In this session we will implement an AI-powered emotion recognition system, and take it all the way from inception, through training a deep learning model and up to deploying a scalable service in production that identifies human emotions in images.

During the session we will review the problem of emotion recognition, we will build and train a neural network to model a solution using Apache MXNet, and finally we will deploy our model to production using Model Server for Apache MXNet and Amazon ECS. Attendees will go through a hands-on overview of an end-to-end development lifecycle for deep learning, which they can later apply to other interesting and challenging problems that can be solved with artificial intelligence.

Agenda:
6:30 PM check-in
7:00 tech talk
8:00 Q&A
8:30 Networking

Speakers

• Hagay Lupesko
  Hagay has been busy building software for the past 15 years, and still enjoys every bit of it (literally)! He engineered and shipped products across various domains: from 3D cardiac imaging, through semi-conductors systems that measures objects the size of molecules, and up to media streaming systems that scale to millions of users world-wide. He is currently an engineering leader at Amazon AI, working on deep learning systems.

-Sandeep Krishnamurthy
Sandeep, software engineer with Amazon AI, is on a mission focused on building tools and technologies to enable usage of AI by millions of developers. He thinks his mission will be complete when developers will use AI just like they use File IO today! Recently, he is busy contributing to Apache MXNet and Keras-MXNet open source deep learning frameworks. In the past, he has built large scale machine learning and data processing platforms at Amazon.