Automatic generation of facial images using EEG and AI

Important results from the project

The goal of the project is to develop the world’s first identikit solution based on AI and brain-computer interface and evaluate the proof-of-concept system on participants from the outside of the lab environment. Both goals were successfully fulfilled. We developed the full system and tested it on 15 volunteers who signed up for participation based on our public ad. We could not involve the Swedish Police in our project because of COVID and the lack of resources at the Swedish Police for the period the project was running.

Expected long term effects

Technology: We successfully developed and tested the BCI-based automated identikit solution. Our results show that we can produce accurate mental images of human faces using our system. We developed the MVP for the criminal investigation’s application. Upon successful collaboration with the Police, we expect to release a commercial version of our solutions in the near future. Scientific outcome: We are now preparing a scientific paper describing our findings to be submitted to top-tier conferences on human-computer interfaces.

Approach and implementation

Technology: We developed the prototype of the product and evaluated it with volunteers without prior experience with BCI. To release the product, we should improve the scalability to generate wider range of face images and optimize the hardware to make it portable and cheap. We plan to build an engineering team further develop it. Market: We plan to hire a business developer to assess new market opportunities. Financing: We plan to apply for financing to grow our engineering and business teams.