Electronic Arts has been granted a patent for a method to train two reinforcement-learning agents in a computer game environment. The first agent generates sub-goals based on the second agent’s performance, with rewards issued for successful achievements. Once trained, the agents can automatically generate content and interact within the game environment. GlobalData’s report on Electronic Arts gives a 360-degree view of the company including its patenting strategy. Buy the report here.

According to GlobalData’s company profile on Electronic Arts, Location-based parallel gaming was a key innovation area identified from patents. Electronic Arts's grant share as of January 2024 was 72%. Grant share is based on the ratio of number of grants to total number of patents.

Training rl agents for procedural content generation in computer games

Source: United States Patent and Trademark Office (USPTO). Credit: Electronic Arts Inc

A recently granted patent (Publication Number: US11883746B2) outlines a computer-implemented method for training two reinforcement-learning (RL) agents within a computer game environment. The method involves iteratively generating sub-goals by the first RL agent, updating based on rewards issued when the second RL agent achieves these sub-goals, and outputting final agents for procedural content generation and interaction within the game environment. The second RL agent iteratively interacts with the game environment to achieve sub-goals, updating based on rewards issued by the environment. The patent also includes provisions for applying auxiliary diversity signals to reward functions, utilizing a population of second RL agents, and employing neural networks for the agents.

Furthermore, the patent describes a generator RL apparatus for procedural content generation in a video game environment and a system for training the first and second RL agents. The system includes modules for generating sub-goal sequences, interacting with the game environment, updating the agents based on rewards, and outputting final agents for automatic content generation and interaction. The method and apparatus detailed in the patent aim to enhance the training and performance of RL agents within computer game environments, ultimately improving procedural content generation and player interactions within the gaming experience. The use of RL techniques, neural networks, and iterative goal achievement processes are central to the innovation outlined in the patent, offering a structured approach to training agents for optimal performance in gaming scenarios.

To know more about GlobalData’s detailed insights on Electronic Arts, buy the report here.

Premium Insights

From

The gold standard of business intelligence.

Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.

GlobalData

GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.

GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.