Research in alternative HCI Methods
MindSki: Just Focus.
Category
Game Development
Status
Completed
Year
2024
Method
Web Technologies
Overview
The idea of this game came up to me when I finished developing Claw which is quite rich in its philosophy and in some ways I can describe it as introducing technology to the realm of art; so I started to think about inviting arts into the realm of technology, in other words bring art to our practical usages, and the idea slowly developed to MindSki.
What this game does is dynamically increase or decrease the difficulty of the game with a designed algorithm which compares the user’s current attention level with a set Aim level, (currently set to 70), so that it should be able to keep the user’s attention level around the Aim Level. Higher score means that the user has a better ability to concentrate. The parameters are adjusted with several dozens of experiments on myself so that it delivers the best performance to help a user change their attention level.
This game is developed using Three.js as renderer, Nuxt.js as website framework and a Sichiray Taurus (similar to Neurosky Mindwave Mobile) as EEG device. It also features a pencil line shader written with GLSL to make the scene look better.
Feature:
Pencil Line Shader
A customized Pencil Line Shader is written with GLSL based on Sobel operators.
Feature:
Hit The Tree
Three steps are executed after the user hits a tree,
1. One life is lost (10 lives in total)
2. Tree is removed to avoid repeated collision;
3. Window shakes to provide a strong visual stimulus to inform users about the hit.
Feature:
EEG Connection
The Game integrates with an EEG Device (Sichiray Taurus) to get the real-time attention level of
Quantitative Analysis
After a game I would always put my data into python and get a few more insights on how me and the game are performing.
The above diagram shows an example set of attention level and analysis obtained from the game.
The blue line shows the real-time attention level throughout the game, but you can see that it’s not very stable and there are outliers, so the actual value processed by the game is the orange one, which averages the value of previous 3 seconds. 70 is our Aim Level and I’ve created an index to show how close the attention level is from the Aim Level, the lower the better.
It shows a trend that the user’s attention level is gradually closer to the aim level, in other words, Mindski is proved to be working.
More Information
1. The project is playable at: https://ski.qwerzl.me.
2. The project is opensourced at: https://github.com/qwerzl/game.