Rundown:
- Google has recently published a patent for a system for training an actor component to execute real-time gameplay actions in a video game.
- The system uses one or more gameplay data models generated by a remote learning service to improve the artificial action gameplay of the actor component.
- The gameplay data model is modified based on observational data generated from gameplay actions performed by both the local actor component and a human player.
- The modification of the gameplay data model is triggered by meeting predefined criteria, such as a defined duration period, a defined quantity of observational data, and explicit requests received at the remote client computing device.
- The patent aims to address the problem of scaling video game testing processes to match the complexity of modern video games.
Google has recently been granted a patent titled, “EFFICIENT GAMEPLAY TRAINING FOR ARTIFICIAL INTELLIGENCE,” filed in April 2022 under GOOGLE LLC. The patent, published last week, outlines a system for training an actor component to execute real-time gameplay actions in a video game based on one or more gameplay data models generated by a remote learning service. Simply put, the method is to teach a character in a particular video game how to make decisions and take actions based on data gathered from the video game. The character learns from a remote service and can make decisions in real time during gameplay.
The learning service generates a gameplay model, and the character performs actions based on it. The service then uses the results to modify the model and send it back to the character, improving its actions. The system may also use observational data generated by a human player to modify the model further.
“Systems and methods are described for training a locally executed actor component to execute real-time gameplay actions in a gaming application based on one or more gameplay data models generated by a remote learning service. A gameplay data model for the gaming application is provided from one or more server computing systems executing the remote learning service to the client computing device,” reads the abstract for the patent. “Observational data is generated by the local actor component based on in-game results of artificial gameplay actions performed by the local actor component, based at least in part on inferences generated by the actor component using the provided gameplay data model.”
In January, we covered a similar patent from Electronic Arts that describes an artificially intelligent system for automating the quality assurance (QA) testing of video games using imitation learning. It seems like Google aspires to do something similar with this system to address the problem of scaling video game testing processes to match the complexity of modern video games.
According to the patent’s claims, the method involves a server providing a gaming data model to a remote client device, receiving observational data generated from artificial gameplay actions performed within the video game by an actor component executing on the remote client device, modifying the gaming data model based on the received observational data, and provide the modified gaming data model back to the remote client device.
Additionally, the method involves receiving control information associating each output state, such as player reference position, object position, motion vector, geometry information, and in-game reward indicators, of the video game with an input variable of the actor component on the remote client device. Furthermore, each output variable of the actor component is associated with an action available to a human player.
In addition, the gaming data model can be improved by using data from both artificial actions performed by the video game’s system and by observing the actions of human players. The data collected from human players is used to modify the model when a certain amount of data is collected or when a specific request is made. This helps improve the video game’s performance and provides a better experience for the players. “The method of claims 8, wherein the at least on criterion comprises at least one of a defined duration period, a defined quantity of observational data and an explicit request received at the remote client computing device,” elaborates the patent.
It is uncertain if Google plans to incorporate this technology in its video game development procedures. Nevertheless, the potential of this technology is significant for video game companies, as it can streamline the development process, enhance its quality and efficiency, and ultimately reduce the time it takes to release new video games. The specific approach Google will take to introduce this technology to its video game platforms remains unknown, and only time will reveal the answer.
What do you think about this? Do tell us your opinions in the comments below!
Similar Reads: Sony Patents Eye Tracking Controls For MP Narrative Games
Good job! Please give your positive feedback 😏
How could we improve this post? Please Help us. 💡
From writing short stories in his room to finding true enthusiasm for video game and computer hardware journalism, Huzaifa plays video games and write all the latest and greatest news about them. Currently pursuing a Bachelor of Science degree in Data Science, he dives deep into the news, authenticating every tiny detail to serve his audience. When he’s not breaking news, he becomes a master storyteller, conjuring up captivating tales from the depths of his imagination. With a wealth of experience as a Video Game Journalist. He has also worked with Publishers like eXputer, The Nerd Mag and Gamesual making him an expert in Gaming News Industry.
