EA Allowing Players To Recreate Playable Video Clips In Games
About EA | Source: Electronic Arts
Electronic Arts is supposedly working on a technology using artificial intelligence that will enable players to automatically reenact video clips inside video games and even play the re-creations themselves, according to a recently published patent.
Not only will the technology supposedly reenact a certain action or sequence of actions, but it will also supposedly recreate the objects and their placements in the video clip and project them within the environment of the particular video game.
The technology uses a video translation model to determine all the actions and virtual objects within the video clip(s), which can then be mapped onto the actions and virtual objects within the particular video game for reenactment.
The company intents to improve video and/or online gaming for players by enabling them to reenact real-world sequences of actions virtually within a video game.
With commercially and critically successful franchises like Battlefield and FIFA under its name, it’s no secret that Electronic Arts is one of the most widely recognised video game companies in the world. Having released some of the most cherished video games among players, like It Takes Two and The Sims 4, the company is constantly trying to improve its video game franchise for better accessibility and playability of its video games. These improvements from the company come in the form of patents filed and published by Electronic Arts, which discuss innovative and sometimes revolutionary technologies that the company may be working on for its video game franchises.
In August 2021, the company announced its Accessibility Patent Pledge, which enabled all developers to use any of the company’s accessibility-centred technologies. Since then, many patents have come forward from Electronic Arts that are focused on improving both the accessibility and playability of its video games. One such patent from the company discusses a technology that will supposedly use artificial intelligence to enable players to automatically reenact video clips inside video games and even play the re-creations themselves.
Earlier today, we came across a recently published patent from Electronic Arts under the name of Electronic Arts Inc. titled “Interactive reenactment within a video game,” which was filed in August 2021 but published earlier this week. The abstract for the patent reads, “A video reenactment system and method analyze a video clip that a video game player wishes to reenact and maps objects and actions within the video clip to virtual objects and virtual actions within the video game. A reenactment script indicating a sequence of virtual objects and virtual actions as mapped to objects and actions in the video clip is generated using a video translation model and stored for use in reenacting the video clip.”
It further reads, “The reenactment script can be used within the video game to reenact the objects and actions of the video clip. The reenactment of the video clip may be interactive, where a player may assume control within the reenactment and when the player relinquishes control, the reenactment will continue at an appropriate part of the sequence of actions by skipping actions corresponding to the ones played by the player.” While this technology may seem like pseudoscience, it seems like Electronic Arts has actually found a way to implement it using artificial intelligence.
As the company explains in the patent, the intention behind this technology is to allow players to replicate an action or sequence of actions performed in a particular video clip within a video game without the requirement of any prior skills for the re-creation, as is usually necessary for complicated reenactments. For instance, recreating a RendeZook in Battlefield 2042 may no longer be all that difficult, as artificial intelligence could seemingly be able to recreate that automatically without any endeavour from the player just by analysing a video clip with the manoeuvre. Eventually, the patent aspires to enable players that don’t have enough skills as are required for a certain action or sequence of actions to be performed within a video game to recreate their desired instance with only a video clip.
However, this isn’t all. Not only will this technology supposedly reenact a certain action or sequence of actions, but it will also supposedly recreate the objects and their placements in the video clip and project them within the environment of the particular video game. “The reenactment may involve virtual objects (e.g., avatars, weapons, buildings, etc.) in the video game that are similar to the objects in the video clip performing virtual actions (e.g., movements, carrying other objects, explosions, etc.) within the video game that mimic the actions in the video clip,” the patent reads. “Thus, the mechanism for reenacting video clips in a video game, as disclosed herein, involves identifying objects in the video clip and mapping them to objects or assets in the video game, and further, mapping actions in the video clip to actions in the video game.”
Furthermore, the patent also mentions that the video clip(s) can be from any suitable source, such as a live-streaming platform or social media platform, or even personally recorded video clips, which will then be analysed using a “video translation model” that will determine all the actions and virtual objects within the video clip(s). As such, the determined actions and virtual objects will then be mapped onto the actions and virtual objects within the particular video game for reenactment. Hence, players definitely won’t be able to shoot a weapon inside FIFA 23 (obviously).
As an example of the reenactment, the patent mentions NBA LIVE 19, where “a player may wish to reenact a sequence of actions from a basketball video clip within a basketball video game.” The patent further explains, “In the basketball video clip, a first player may rebound a ball on the defensive end and make a long pass to another player at the offensive end who dunks the basketball.” Subsequently, the basketball video clip would then be analysed by the video translation model associated with the particular basketball video game. “Using the video translation model, a first object (e.g., the first player in the basketball video clip) may be mapped to a first virtual object (e.g., a virtual first player) in the basketball video game. Similarly, the second player in the basketball video clip may be mapped to a second virtual player in the basketball video game. Other objects may also be identified, such as other players, the basketball, the coach on the sideline, etc,” it reads. Eventually, the reenactment script may then be used to reenact the sequence of objects and actions within the basketball video clip.
Additionally, the reenactment script associated with a particular video game may contain not only information about the actions and virtual objects, like their placement, but also other minor details, like which way non-player characters (NPC) are faced or even the velocity associated with the movement of the virtual objects. In addition to this, the reenactment may also supposedly be interactive, allowing players actually to control and play the re-creations themselves. “Thus, a player may commence a sequence of actions of a reenactment and at some point within the reenactment may wish to take over in controlling the actions within the video game prior to the conclusion of all of the sequence of actions of the reenactment,” reads the patent. The reenactment could also allow players, if they desire, to revert back to a certain instance within their re-creations to play it out differently.
Lastly, the company mentions the intention behind this technology is to not only “improve video and/or online gaming for players, via their player accounts, by enabling players to reenact real-world sequences of actions virtually within a video game” but also to improve “a variety of other technical fields related to video gaming, such as virtual reality, augmented reality, etc.” While it’s uncertain exactly how this technology will be implemented within the company’s video game franchises, if it all, there may still be some concerns from players, as many of the reenactments and re-creations within video games require considerable skillfulness and experience, which would be made almost obsolete with artificial intelligence doing all the exertion instead of players.
For instance, difficult trick shots in a video game franchise like Battlefield may no longer be a cause for celebration or amazement, as artificial intelligence may be able to re-create them automatically without the need for any experience or endeavour from the players themselves. However, there’s no denying that this technology will certainly revolutionise the way reenactments and re-creations are made within video games, and only time will tell exactly how this technology will come to fruition.
What do you think about this? Do tell us your opinions in the comments below!