The project is an implementation on the paper :

  @misc{pang2021predictive,
  title={Predictive coding feedback results in perceived illusory contours in a recurrent neural network}, 
  author={Zhaoyang Pang and Callum Biggs O'May and Bhavin Choksi and Rufin VanRullen},
  year={2021},
  eprint={2102.01955},
  archivePrefix={arXiv},
  primaryClass={cs.CV}
  }

The team members involved in this project are:

 Sagar B Dollin, MSc Cognitive Science, IIT Delhi
 Adheena Reji, MSc Cognitive Science, IIT Delhi
 Akshat Govekar, MSc Cognitive Science, IIT Delhi

This paper gives a method to make computer vision to perceive Kanizsa illusion. The method used is Predictive Coding Feedback Recurrent Neural Network that was proposed by Rao and Ballard. Please read the provided Report to understand the implementation. For a more broader intuitive understanding refer to "The philosophy and science of illusion" file present in the same repository to explore more on this problem statement.

NOTE: At the time of doing this project the code for the model proposed by the paper we are referring was not released yet, and still is not released (at the time of uploding this project) , The code built is completely built by the author of this directory, Sagar B Dollin using some refrences like : https://youtu.be/JoR5HCs0n0s and https://youtu.be/7HPwo4wnJeA etc.