• Artificial intelligence is learning to read your mind and display what it sees

 

 

• Be able to reconstruct or recreate an image a person is seeing or and even thinking about,  just by looking at the person’s brain activity

 

 

What is it about?

 

Deep Neural Network (DNN) 

 

 

 

 

  Neural Network:

• mimic the human brain through a set of algorithms
• design to recognise patterns
• rely on training data to learn and improve their accuracy over time
• allow us to classify and cluster data at a high velocity

 

This ML

images

DATA

individual

fMRI

brain activitydata

direct visualization

memorized visualization

*over a ten-month period

*images (geometric shapes, natural phenomena and letters of the alphabet)

individual

Deep Image Reconstruction -  a method to visualize perceptual content from human brain activity

 

Combined the DNN feature decoding from fMRI signals and the methods for image generation.

 

    - built a model of how the brain encodes information

 

    - train the model to associate images with brain activity

 

    - the network partially reconstruct what a participant saw,            decoding brain activity into pixels

https://youtu.be/jsp1KaM-avU

THE MODEL

"The results suggest that hierarchical visual information in the brain can be effectively combined to reconstruct perceptual and subjective images"

 

 A way to access, explore and share the content of our perception, memory, and imagination

 

RESULTS

 

 

- patients in permanent vegetative states to communicate

 

     - make video recordings of daydreams

 

      - therapy (analysis of thoughts)

 

      - criminal investigations

 

 

 

APPLICATIONS

 

• Access to complex and intimate places on our minds

• Be able to know what others might be thinking, can be dangerous...

 

 

 

ETHICAL implications

• Use of this image processing method to associate video images with brain activity and predict what the person was watching (categorize within 15 different objects)

 

• University of Toronto have created an AI that is  is able to recreate images seen by humans based on their brain activity, as collected using electroencephalogram (EEG) sensors.

 

 

 

 

Other similar projects