How Researchers Are Using Machine Learning To Make VR Sword Fights More Realistic
Researchers at the University of Bath in Somerset, United Kingdom, have partnered with Cambridge-based game studio Ninja Theory (Hellblade VR, DmC: Devil May Cry, Enslaved: Odyssey to the West) to deliver more realistic, data-driven sword fighting in VR.
While pre-recorded attacks and defensive positions may work fine in the standard 2D gaming format, the same cannot always be said for VR. Many of the current VR sword fighting experiences rely heavily on this formulaic approach, often requiring the player to memorize a set pattern of moves and time their attacks accordingly.
“With screen-based video games, a player presses ‘attack’ and their character displays a sequence of animations,” said Dr. Christof Lutteroth, Senior Lecturer in Computer Science at the University of Bath. “But in a VR game, the player input is much harder to process.”
These repeated actions can easily break player immersion, which is why researchers have begun working on a new system that uses machine learning to deliver more realistic sword fights against less predictable opponents. Referred to by the team as Touché, the programmable system is powered using a collection of pre-recorded sword-fighting animations captured using motion capture technology.
Using a form of artificial intelligence referred to as “machine learning,” non-playable characters can actually learn from these recorded moves and increase their skill level. Opponents can also be trained to respond more realistically to the attacks and positioning of the human player, resulting in tougher, more complicated engagements. According to researchers, developers can control each enemies “skill and aggression,” allowing them to scale their experience to meet the needs of any player, regardless of skill.
“Touche increases the realism of a sword fight by generating responsive animations against attacks and eliminating non-reactive behavior from characters,” added Dr. Lutteroth. “Using our model, a game character can anticipate all possible fight situations and react to them, resulting in a more enjoyable and immersive game experience.”
While games such as Blade & Sorcery have come close to offering realistic VR sword fights, there’s still a sense of automaticity in the movement of enemy characters. After just several hours in-game, even those unfamiliar with VR will find themselves dancing around enemies with relative ease. Hopefully, by incorporating machine learning into the development process, we’ll soon see a new standard for VR sword fighting powered by human movement.
For more information on Touché, visit here.
Image Credit: University of Bath / Ninja Theory