Table of Contents
Facebook developing is a new system powered by machine learning that will allow individuals to re-create text style from a single word example, resulting in a new translation, editing, and replication opportunities.
With Facebook’s Text Style Brush project, users would be able to replace existing text in photographs while also replacing background images by using just a single photo and just a single word as a reference point.
As noted, this is a slight advance over existing text replication methods and could have a variety of useful applications.
As Explained by Facebook:
“Unlike most AI systems that can do this for well-defined, specialized tasks, Text
Style Brush is the first self-supervised AI model that replaces text in images of both handwriting and scenes – in one shot – using a single example word.”
As Facebook mentions, this project could provide users with the ability to translate text within images and interpret handwritten signs in different languages (see example below) and create personalized messages and captions, among other things.
However, it can also be misused. This technology allows people to easily erase watermarks from images and change the meaning of its elements or even to change the composition of the photograph.
What if, for example, an old college photo of a politician holding a sign was modified so that the candidate seemed to be espousing an offensive message instead of the actual message?
The company is well aware of these risks, as demonstrated in its research paper, which aims to “prompt additional research and dialogue” to help prevent deep fake text attacks.
The Journal Said:
“If AI researchers and practitioners can get ahead of adversaries in building this technology, we can learn to better detect this new style of deep fakes and build robust systems to combat them.”
Although it seems risky, it is still worth a shot. It’s useful, especially if you want to recreate a text style and for evolving translations.
There is, however, the possibility that, at least in terms of broadly publicizing the program, the risks would trump the benefits.
Facebook developing may only ever use it as a translation tool; however, there is some value to it that Facebook will certainly want to capitalize on. This is a fascinating project in either case.