FIFA 22: with HyperMotion, machine learning takes over in animation

FIFA 22: with HyperMotion, machine learning takes over in animation

Spread the love

FIFA 22: with HyperMotion, machine learning takes over in animation

FIFA 22 represents an important technological evolution not only for EA’s football series but also for the way in which video games – sports, but not only – could be developed in the future, given that the introduction of technology HyperMotion determines an important beginning for the machine learning applied to the construction of animations.

As explained by Sam Rivera of EA Sports, FIFA 22 begins to introduce the intensive use of machine learning on recording standard animations, which could involve a radical change in the way these are created in the world of video games.

“We started working on a algorithm about three years ago, “said Rivera, FIFA 22 lead gameplay producer at EA Vancouver,” What this algorithm does is learn from all the data contained in the motion capture recordings – how the player behaves with the ball, how many steps must be taken to approach this, what is the right angle of the body, the rhythm of the movements, how to hit the ball properly “.

All this data is processed by the machine learning algorithm and then applied to the animations of the players.

FIFA 22 is expected to start presenting next-generation animations with HyperMotion and machine learning

All this “Create this solution, with animations in real time. This is a very advanced technology, it is basically the beginning of machine learning that supplants classic animation “.

This has several advantages, because so far much of the work on animations depends on motion capture recording, which is a long and expensive process and can hardly reach a degree of complete completeness. “With access to HyperMotion, we can insert longer animations, and more complex animations for ball control,” explained Rivera, “but this technology also allows us to process some new transitions in the midst of the animations, finding new ones when the situation changes “.

In short, the use of machine learning, and HyperMotion in this case, should guarantee greater differentiation in movements, which should be more appropriate to each situation, adapting in real time. Furthermore, on a broader level, it should guarantee a more sensible and orderly movement of the players on the field, following the general indications on the form and formation.

In the meantime, we’ve seen player ratings revealed, including Messi, Lewandowski, Ronaldo, Donnarumma and Mbappé, for FIFA 22, as well as the new Icons revealed with a trailer.

s.parentNode.insertBefore(t,s)}(window, document,’script’,
fbq(‘init’, ‘1624156791000312’);
fbq(‘track’, ‘PageView’);