9th Deep Learning Meetup in Vienna

Hi Deep Learners,

The main topic of our 2nd meetup this year is "Face Recognition".

Philipp Omenitsch, Research Scientist at VisionLabs (http://www.visionlabs.ru/en/), will show how they made very deep CNNs ready for market applications. Then we are going to continue the "Coding Deep Learning in Keras" tutorial, which was perceived as very helpful to many people last time. This time, Alexander Schindler will show us how to use Recurrent Neural Networks on hard disk data to predict hard disk failures.

Agenda:

• Welcome

• Visionlabs: Face Recognition for Businesses (Philipp Omenitsch)

Seen as a benchmark for deep learning methods in recent years, face recognition has become production ready. Visionlabs (http://www.visionlabs.ru/en/) shows how it brings state of the art research to the market. Visionlabs tries to harness the improvement in accuracy of new CNN architectures which tend to be 1000 of layers deep and include significant computation, while still being fast enough for large scale face recognition for companies with millions of customers.

• Latest News and Hot Topics (Tom Lidy and Jan Schlüter)

• Coding in Keras: Hard-Disk Failure Prediction with SMART data using RNNs (Alex Schindler, Research Scientist at TU Wien and AIT)

• Questions & Discussion

This meetup is thankfully hosted at BWin who will provide us with snacks!

looking forward,

Tom & Jan