Market Sector

This project falls into the entertaining sector.

According to the statistics, in 2020 the global entertainment and media market is expected to value $2.0 trillion, and is expected to keep rising in the coming years. In the U.S., this value was expected as $720.38 billion in 2016.

According to this article, the global video game revenue is expected to be $179.7 billion in 2020.

Main Companies

According to this rank, the largest video game companies in 2020 are Sony ($20.3 billion), Nintendo ($11.1 billion), Microsoft ($11.0 billion), Tencent ($9.2 billion) and Activision Blizzard ($6.5 billion).

Among video game companies invest heavily into animations, Ubisoft interests me the most. The page of La Forge Project shows their efforts to import academic research results into games. The project contains many topics on introducing machine learning or reinforcement learning into motion matching and character controlling.

Products

The products related to this project are mainly video games. It may also bring benefit to the making of 3D animation films. 3D modeling tools and game engines can integrate the animation generator, too. Also it can serve as an alternative to motion capture studios.

  • Cited from Game Anim, motion matching techniques have been applied in more than 5 projects of Ubisoft including Assassin’s Creed III
  • According to this article, Naughty Dog has used motion matching in The Last of Us: Part II
  • Also in Uncharted, characters have physical interaction by using motor controllers following motion capture clips
  • Wolfire Games’ Overgrowth uses only procedurally generated animations without any key-framing or motion capture data, however, these animations are purely kinematic (no physical interaction), though they’ve tried their best to make them look physically correct
  • Videocult’s Rain World uses procedural animations to make character movements fluent and interactive with the environment, which is also purely kinematic

Stakeholders

Game companies and studios making 3D games will benefit mostly on this project, especially for small game studios. For large companies, though motion capture is available to them, it is still not easy to be applied on non-humanoid characters. For small game studios, this technique can save their games from low quality animations and complex animation controllers.

In terms of individuals working in those companies, the project will affect animation programmers the most. Making complicated animation state machines will be a past. The design of character controller won’t be ever more intuitive, since all animation staff is totally decoupled from it.

It’s still not clear to me how it will change the life of animators. Not all animations can be generated, and key-framing is still needed for stylized animations. Maybe it can release them from key-framing animations that must look physically correct (which are sometimes more difficult).