Romain Vuillemot bio photo

Romain Vuillemot

Assistant Professor
École Centrale de Lyon
LIRIS Laboratory

Email Twitter Facebook Google+ LinkedIn Github Youtube Google Scholar

What if we Reduce the Memory of an Artificial Doom Player?

Theo Jaunet, Romain Vuillemot, Christian Wolf

Conference: VISxAI 2019 at IEEE Vis 2019
Proceedings: VISxAI 2019 at IEEE Vis 2019
Demo: https://theo-jaunet.github.io/MemoryReduction/

We built Doom player AI using Deep Reinforcement learning. While playing, it builds and updates an inner representation (memory) of the game, its environment. Reducing this memory could help the player learning to complete its task and thus lower both its training time and energy consumption footprint.

Abstract

We built a Doom player AI using Deep Reinforcement learning. While playing, it builds and updates an inner representation (memory) of what it sees from the game. This memory represents what the AI knows about the game, and is the root of each decision. Reducing the size of the memory , could help the player learning to complete its task and thus lower its training time and energy consumption footprint. In this scenario, the player has to gather items in a specific order: Green Armor Red Armor Health Pack Soul-sphere , with the shortest path possible.


← Back to publications