20240017175. Adversarial Reinforcement Learning for Procedural Content Generation and Improved Generalization simplified abstract (Electronic Arts Inc.)

From WikiPatents
Jump to navigation Jump to search

Adversarial Reinforcement Learning for Procedural Content Generation and Improved Generalization

Organization Name

Electronic Arts Inc.

Inventor(s)

[[:Category:Linus Mathias Gissl�n of Enskede gard (SE)|Linus Mathias Gissl�n of Enskede gard (SE)]][[Category:Linus Mathias Gissl�n of Enskede gard (SE)]]

Andrew John Eakins of Guildford (GB)

Adversarial Reinforcement Learning for Procedural Content Generation and Improved Generalization - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240017175 titled 'Adversarial Reinforcement Learning for Procedural Content Generation and Improved Generalization

Simplified Explanation

Methods, apparatus, and systems are provided for training two reinforcement-learning (RL) agents in a computer game environment using RL techniques. The first RL agent generates a sub-goal sequence in relation to an overall goal within the game environment. After the second RL agent successfully achieves a current sub-goal in the sequence, the first RL agent generates a new sub-goal. The second RL agent iteratively interacts with the game environment to achieve the current sub-goal. The first RL agent is updated with a reward when the second RL agent successfully achieves the current sub-goal, and the second RL agent is updated based on a reward issued by the game environment. Once trained, the first RL agent can automatically generate procedural content in the game environment, and the second RL agent can automatically interact with a procedural content generation (PCG) game environment.

  • The patent describes a method for training two RL agents in a computer game environment.
  • The first RL agent generates a sub-goal sequence in relation to an overall goal in the game environment.
  • The second RL agent interacts with the game environment to achieve the current sub-goal in the sequence.
  • The first RL agent is updated with a reward when the second RL agent successfully achieves the current sub-goal.
  • The second RL agent is updated based on a reward issued by the game environment.
  • Once trained, the first RL agent can automatically generate procedural content in the game environment.
  • The second RL agent can automatically interact with a PCG game environment.

Potential Applications

  • Automatic procedural content generation in computer games.
  • Training RL agents to interact with game environments.

Problems Solved

  • Manual generation of procedural content in computer games.
  • Manual interaction with game environments by RL agents.

Benefits

  • Improved efficiency and speed in generating procedural content.
  • Enhanced gameplay experience through automated RL agent interaction.


Original Abstract Submitted

methods, apparatus and systems are provided for training a first reinforcement-learning (rl) agent and a second rl agent coupled to a computer game environment using rl techniques. the first rl agent iteratively generates a sub-goal sequence in relation to an overall goal within the computer game environment, where the first rl agent generates a new sub-goal for the sub-goal sequence after a second rl agent, interacting with the computer game environment, successfully achieves a current sub-goal in the sub-goal sequence. the second rl agent iteratively interacts with the computer game environment to achieve the current sub-goal in which each iterative interaction includes an attempt by the second rl agent for interacting with the computer game environment to achieve the current sub-goal. the first rl agent is updated using a first reward issued when the second rl agent successfully achieves the current sub-goal. the second rl agent is updated when a second reward is issued by the computer game environment based on the performance of the second rl agent attempting to achieve said current sub-goal. once validly trained, the first rl agent forms a final first rl agent for automatic procedural content generation (pcg) in the computer game environment and the second rl agent forms a final second rl agent for automatically interacting with a pcg computer game environment.

Retrieved from "http://wikipatents.org/index.php?title=20240017175._Adversarial_Reinforcement_Learning_for_Procedural_Content_Generation_and_Improved_Generalization_simplified_abstract_(Electronic_Arts_Inc.)&oldid=28143"