Exploring Deep Recurrent Q-Learning for Navigation in a 3D Environment
Journal Title: EAI Endorsed Transactions on Creative Technologies - Year 2018, Vol 5, Issue 14
Abstract
Learning to navigate in 3D environments from raw sensory input is an important step towards bridging the gap between human players and artificial intelligence in digital games. Recent advances in deep reinforcement learning have seen success in teaching agents to play Atari 2600 games from raw pixel information where the environment is always fully observable by the agent. This is not true for first-person 3D navigation tasks. Instead, the agent is limited by its field of view which limits its ability to make optimal decisions in the environment. This paper explores using a Deep Recurrent Q-Network implementation with a long short-term memory layer for dealing with such tasks by allowing an agent to process recent frames and gain a memory of the environment. An agent was trained in a 3D first-person labyrinth-like environment for 2 million frames. Informal observations indicate that the trained agent navigated in the right direction but was unable to find the target of the environment.
Authors and Affiliations
Rasmus Kongsmar Brejl, Henrik Purwins, Henrik Schoenau-Fog
Keywords
Related Articles
- EP ID EP45886
- DOI http://dx.doi.org/10.4108/eai.16-1-2018.153641
- Views 292
- Downloads 0
Exploring avatar roles for motivational effects in gameful environments
This paper explores avatar roles and design principles in helping to develop motivation in game environments. Different avatar applications could influence the experience of players in at least three ways: as a customisa...
Establishing Interaction between Machine and Medaka using Deep Q-Network
Social interaction is the basic ability for animals to survive. It is difficult for a machine to interact with human or other animals because it is not clear how the machine should interact. This paper examines whether a...
Maze and Mirror Game Design for Increasing Motivation in Studying Science in Elementary School Students
The research project discussed here, examines attempts to increase the motivation of elementary school students in basic science by the means of designing a science game. To realize this goal, the maze and mirror game wa...
Packet Replication and Noise in Reliable End-to-end Molecular Communication
In modern communication systems, noise represents the source of errors that disturb useful information signals for communication. In biological systems, however, noise plays important roles for their functions. In this p...
Solitude or co-existence – or learning-together-apart with digital dialogic technologies for kids with developmental and attention difficulties
An overall political vision of a prosperous society is one in which everyone has the same access and possibilities of participating in democratic processes, and in which everyone has equal access to the resources, life a...