[ad_1]
Researchers from Högskolan i Skövde have "formed" a network of artificial neurons that, thanks to a conventional webcam, can detect emotional states in a person playing a computer game. This can pave the way for a future where computer games will adapt to the player's feelings.
Imagine playing a computer game. While playing movies, a camera on the face of your computer. By analyzing the video, the computer can determine if you think the game is fun, boring, easy or difficult. The game then adjusts to optimize your gaming experience.
The computer games available that nowadays adapt to the real feelings of a player use facial recognition and uncomfortable physical sensors. But Högskolan i Skövde's research can change that. During his studies, the PhD student Fernando Bevilacqua built a network of artificial neurons able to detect the emotional state of a player by analyzing a video of the player's face.
Tested with Super Mario Bros.
A network of artificial neurons is a kind of AI software that tries to mimic the working methods of neurons in the human brain. Fernando Bevilacqua is "trained" and tested his artificial neuron by allowing a number of people to play a modified version of the game Super Mario Bros.
At the same time, a regular webcam filmed each player's face. The video was then subjected to a process called photoplethysmography, in which the player's heart rate can be read using only the video. Information about the player's face expression was also extracted from the video before all the information was entered into the artificial neural network. Using only video information, the neural network could correctly classify the emotional state of a player in more than 60% of cases.
Fernando Bevilacqua presents the software that he used to "train" an artificial neuron. Picture: Högskolan i Skövde
"During my research, I focused on games and used a game to calibrate the artificial neural network, but I think this technology can also be used in other areas. perhaps we could use it in cars to inform the driver when he is in an emotional state where it can be dangerous to drive, "says Fernando Bevilacqua.
Play at your specific level
Today, questionnaires and physiological measurements are used if someone wants to assess the emotional conditions of the player playing a computer game. These methods give searchers and game developers an average of how a group reacts to a game.
"The basic idea of our research is to move away from the group and focus more on each player's experience, which opens the way to a future where you no longer play an easy-to-play game. Medium or difficult, instead you play the game at your own level, "concludes Fernando Bevilacqua.
Contact:
Fernando Bevilacqua, [email protected]
dissertation:
Remote sensing of emotions, calibrated for the game and adapted to the user: non intrusive multifactorial camera approach to detect the stress and the boredom of the game players
Source link
