Tracking down fakes together: The potential of AI

Event from Warwick AI

Wednesday 16 September 2020, 2pm - 3pm


Is seeing believing? When you’re scrolling through Instagram how true to life are the posts you see? 

With the wide-spread use of photoshop to enhance images while simultaneously making those enhancements un-noticeable to the naked eye, those questions become tougher to answer. 


The latest policies show a push against photoshop use. A recent move by the French government instructs the disclosure of photoshop in adverts where bodies have been made artificially thinner or thicker [1]. 

Whilst Adobe has developed an AI to detect the use of photoshop in images, it has collaborated with the New York Times and Twitter on the ‘Content Authenticity Initiative’ [2] ensuring fakes are exposed. 


AI models are driven by data. In this case, a lack of photoshopped images is preventing such algorithms from coming to life and changing how such photoshopped images are policed. 


Whether you can make the algorithms or help collect a huge dataset, we believe the student community will be the key to solving this problem.




Share on social media?

You may also like?