Neuro-Controller

I use Emotiv's EPOC X EEG headset to record brain waves. Using an interface I developed in Python, I connect the headset to the computer and use it as a controller. This allows the headset to be used to control the PC or other devices.

Similar to training artificial intelligence, the Emotiv software can be used to train patterns of brain waves as so-called commands. For each command, a specific idea should be consistently thought of so that the pattern is reliably stored. The EPOC X has 14 channels and can store up to four commands.

With the help of Python code, I can intercept and process the commands trained via the BCI software. In my case, a command triggers the activation of a button on a virtual controller. This makes it possible to control games without using your hands.

The Python code establishes a connection to the Emotiv API and reads the commands triggered by the headset in real time. At the same time, a virtual PC controller is created using Python. The input mapping is currently still hardcoded, as this is not the final version.

Since the headset can only store up to four commands, the number of available inputs is severely limited. Depending on the game, this can make meaningful control impossible. In addition, it is recommended to train for about eight hours per command in order to achieve reliable pattern recognition, which is a significant disadvantage.

In addition, the headset has a slight delay when reading brain waves and recognizing commands. This makes playing real-time games much more difficult.

I enjoy experimenting with this device, even though it has significant limitations compared to a standard controller. However, it shows that it is already possible to play video games without using your hands. Such technology can make a valuable contribution, particularly in the area of accessibility, even if this is only possible to a limited extent with the device used here due to the high setup costs.