#13 - 2018-05 - AI-Bot playing Doom; ColorTransfer - CNN colorization

May 24, 2018 · München, Germany

SLOT1: AI-Bot playing Doom - Applied Deep Reinforcement Learning - Rodolphe Masson

In this presentation, Rodolphe shows the results (videos) of his investigations of various deep reinforcement learning techniques, standard deep Q-network (DQN), Replay Memory (RM), Double DQN and Dueling DQN, applied to a software agent whose goal is to play autonomously the game Doom through three different scenarios of increasing complexity, “basic”, “health gathering supreme” and “deadly corridor”, from the ViZDoomEnvironment.
Then he concludes on the most promising techniques to pursue this project in the future like Asynchronous Advantage Actor-Critic (A3C) and Direct Future Predictions (DFP)

SLOT2: ColorTransfer: Bringing New Life to Old Photos Using Neural Nets - Preslav Rachev

Convolutional neural networks are all the rage nowadays, especially, when it comes to digital imagery. They are equally well suited for image classification, as well as for generating new images, based on certain pre-trained models. Can they then help bring a colourful life back to old black and white photos? Let’s check it out!

Inspired by existing research on photo colorization, I have devised my own set of empirical experiments based around the use of CNNs and autoencoders.

SLOT3: Advanced Analytics and IoT explained with a show case - Florian Haeussinger

The project implements an IoT-Showcase with small toy cars of the brand Anki OVERDRIVE. The vehicles have an IR-Sensor, a microprocessor and an electro motor. Data and commands are transmitted to the vehicles or the controller using Bluetooth Low Energy (BLE).
The collected data is stored in a Lambda-Architecture, which is based on open source software solutions. From there, various scenarios like anti-collision or product-improvement using IoT technologies and machine learning methods were developed.

We are looking for speakers and sponsors!

As always, we are super happy to have you with a lightning talk on your small or big successes, open source projects or just something amazing you would like to share.

Do not miss out updates on twitter - https://twitter.com/hack_ai and join Munich ML community on Slack channel https://ai-hack-inviter.herokuapp.com/

Event organizers
  • Hacking Machine Learning

    All hands-on meetups around hacking traditional machine learning down to creating some intelligent agents and self-driving models. Who should join:  We are open to anybody who is interested in Machine Learning or Data Science from really new, but already passionate up to experienced folks, who would like to play around side projects. Why members should join: To learn, share, and have fun doing so. What members can expect: Monthly meeting with knowledge sharing on related topics Do not miss out updates on

    Recent Events
    More

Are you organizing #13 - 2018-05 - AI-Bot playing Doom; ColorTransfer - CNN colorization ?

Claim the event and start manage its content.

I am the organizer
Social
Rating

based on 0 reviews