chitkara logo


Vol. 3, Issue 46, December 2017
Decoding Human Brain
As technology is burgeoning day-by-day, the areas of research are pushing limits that are unimaginable. Perfect example of this statement is new development by scientists and researchers in the field of Artificial Intelligence (AI), which can decode human mind by scanning brain activity.

Image source: economictimes.indiatimes.com

It is now possible to predict what a person's mind is calculating or thinking at any particular moment. The brain decoding technique is based on neural networks technology to understand the scenario of human mind and uses brain scanning techniques like functional Magnetic Resonance Imaging also known as fMRI, to see which areas of the brain are active during various activities.

Brain decoding looks inside the active areas to identify very fine patterns related to images or ideas. The study involves putting subjects (people) in a scanner and showing them a series of pictures. For each picture, scanner records activity in the areas of the brain responsible for vision. A computer program - 'the decoder' is then trained to associate each picture with its pattern. To test the decoder a new picture is presented to subjects, the scanner records the activity again. Decoder then compares new pattern to the pattern it knows from training to figure out what type of object the person was looking at.

Dr. Jack Gallant of University of California, Berkeley with his team measured brain activity of their subjects as they watched hours of movie footage. They taught a computer program to associate the clips with brain activity patterns. Then the decoder reconstructed what the subject had seen based on the brain activity alone. The computer shifted through lots of video clips and produced a best guess that it thought was most similar to what the subject had seen.

Dr. Yukiyasu Kamitani of Kyoto University, Japan with his team carried out a study where they made people to fall asleep in a brain scanner and then wake them up periodically and asked them what they were dreaming about. They noted down the most popular categories of dream objects. Then by using the machine learning algorithms and fMRI imagery they found patterns to predict the dreams of the people using their device which they call 'dream-trained decoder'.

This technique of human brain decoding using Artificial Intelligence could be used to extract secret thoughts and even to detect lies. Scientists are also presuming that this technique could help in finding the hidden facts about activities happened in past with humans.

By - Ms. Monica, Assistant professor CSE, Chitkara University, H.P.

References:-

  1. https://www.sciencedaily.com
  2. www.thehindubusinessline.com
  3. ddnews.gov.in

About Technology Connect

Aim of this weekly newsletter is to share with students & faculty the latest developments, technologies, updates in the field Electronics & Computer Science and there by promoting knowledge sharing. All our readers are welcome to contribute content to Technology Connect. Just drop an email to the editor. The first Volume of Technology Connect featured 21 Issues published between June 2015 and December 2015. The second Volume of Technology Connect featured 46 Issues published between January 2016 and December 2016. This is Volume 3.

Previous Issue



Nanophotonics in Electronics & Communication
Click here!

Archives - Random Issue from Vol. 1 & 2



Machine Learning
Click here!

Editorial Team

Chief Editor: Sagar Juneja
Members: Gitesh Khurani,
Arun Goyal.

Disclaimer:The content of this newsletter is contributed by Chitkara University faculty & taken from resources that are believed to be reliable.The content is verified by editorial team to best of its accuracy but editorial team denies any ownership pertaining to validation of the source & accuracy of the content. The objective of the newsletter is only limited to spread awareness among faculty & students about technology and not to impose or influence decision of individuals.