Generative Vision in AI

🚀We are excited to announce our next event in the series of Generative Vision in AI meetups. This time we partnered up with TomTom (https://www.tomtom.com/de_de/), leading company in easy-to-use maps, navigation software, and real-time traffic information and services.

Our speaker is Nikolaas Steenbergen, Senior Software Engineer at TomTom, and Brighter's Deep Learning Researcher Ignacio Alvizu will moderate the panel discussion – TBA.

Meet-up Agenda:

6:00 PM – 6:30 PM – Open doors, networking – drinks (sponsored by TomTom) 🍾🍷🥤

6:30 PM – 6:40 PM – Welcoming words – TomTom

6:40 PM – 7:05 PM – talk: 📣Nikolaas Steenbergen – Normalizing Flows for Domain Adaptation

7:05 PM – 7:15 PM – Q&A

7:15 PM – 7:20 PM – Short intro Brighter AI & Meetup

7:20 PM – 7:45 PM – Ignacio Alvisu – Panel Discussion – Challenges of generative models

7:45 PM – 8:00 PM – Q&A

8:00 PM – Onwards – Networking

------------------------------------
Abstract topics:

Normalizing Flows for Domain Adaptation –

Deep learning requires a lot of annotated training data. Data is abundant, unfortunately, annotations are sparse. One potential solution is Domain Adaptation: Finding a transformation function to convert artificial, easy to label data, to realistic looking data. Generative Neural networks can be a good fit to learn this transformation function. Normalizing Flows are fully invertible generative probabilistic networks with advantages over the more popular Generative Adversarial Networks and Variational Autoencoders. We will see how Normalizing Flows work, and how they could be applied to Domain Adaptation.

Panel Discussion – Challenges of generative models

Bio

Nikolaas Steenbergen – Senior Software Developer for Artificial Intelligence and Autonomous Driving at TomTom in Berlin.

As part of the TomTom team, he is working on automating the map creation process from raw input sensor data using machine learning to provide millions of users with detailed and accurate maps.
In his master thesis for the masters Artificial Intelligence at the University of Amsterdam, he extracted chord information from musical signals with deep learning. During his work at the Research Center for Artificial Intelligence, he gathered extensive experience in applying algorithms at large scale on distributed systems.
As a Ph.D. candidate at the University of Delft, his research focuses on Invertible Neural Networks.

Ignacio Alvizú – Deep Learning Engineer at Brighter AI.

As part of the Brighter AI development team, he is working on the development of Brighter’s Deep Natural Anonymization by defining and applying deep neural networks as well as contributing to the software development.
He received his physics master’s degree at Pontificia Universidad Católica de Chile where he studied the quantum interactions between point defects in diamond and their environment. Point defects in diamond, particularly the nitrogen-vacancy (NV) center, have novel applications in quantum cryptography, nanoscale magnetic sensing, and biophysics, among others; being good candidates for quantum bits units.

How to arrive at TomTom:
From S-Bahn Treptower Park, take the exit away from the park.
Cross street, turn right, go straight uphill and enter the higher tower.
Register at the entrance, meetup is just straight on the ground floor.

🔗💻🚀🌏Looking forward to seeing you all!👩👨🧑🎉

Ps. If you are interested in joining us as speakers or sponsors, please get in touch, we are always looking for new and exciting collaborations in the Generative AI space!