Mindfulness Mirror

James-Reynolds, Carl and Currie, Edward (2018) Mindfulness Mirror. In: BCS Machine Intelligence competition, 05 Oct 2018, BCS London Office, First Floor, The Davidson Building, 5 Southampton Street, London.

[img] Microsoft PowerPoint (13th BCS Machine Intelligence entry) - Other
Restricted to Repository staff and depositor only

Download (2MB) |

Abstract

The Mindfulness mirror was developed from a simple infinity mirror that changes its hues in accordance with relaxation levels read from a Neurosky Mindwave EEG. The new mirror uses an interactive Genetic
Algorithm with fitness ratings provided from the EEG, consisting of the sum of attention and meditation (mindfulness) readings. The software runs on the Arduino platform. There are a number of aspects
of the system that could be improved, but having reached the limits of the available memory, this would require porting to a Teensy platform with more memory and faster libraries for managing the 202 LEDs. The system might also be more effective if the mutation rate was slowly reduced over the generations, but these are issues to be explored in future work. Whilst more LEDs would allow geometry to be explored as a variable, this is problematic as it would require a much larger power supply than the 10 Amps we currently use.

The system was developed for the 13th BCS Machine Intelligence Competition; the current presentation was developed specifically for this and lasts about 15 minutes. To demonstrate effectively, we use
only 4 parents, employ a high mutation rate (5%) and limit the display time of each parent to 30 seconds. Co-dominance is used, although a splicing function is also available. We do normally see some convergence over this time period. The project was developed for fun and exploration of ideas but it does raise interesting issues with regard to sensors providing fitness functions and the lack of a clearly defined endpoint. A library will be available for the Arduino platform over the next few months, which should be useful as a teaching tool. We would also like to explore possible applications of the technology for art installations and therapeutic systems.

The talk itself is silent!

Item Type: Conference or Workshop Item (Other)
Research Areas: A. > School of Science and Technology > Computer Science > Artificial Intelligence group
Item ID: 25521
Useful Links:
Depositing User: Carl James-Reynolds
Date Deposited: 05 Nov 2018 11:57
Last Modified: 05 Nov 2018 11:57
URI: https://eprints.mdx.ac.uk/id/eprint/25521

Actions (login required)

Edit Item Edit Item

Full text downloads (NB count will be zero if no full text documents are attached to the record)

Downloads per month over the past year