This algorithm perfectly imitates your calligraphy

Scientists from the University of London (UCL) were able to develop a program that is capable of analyzing the calligraphy of an individual and thus generate semi-automatically a text that could easily have been written by the same person. It turns out to have amazing accuracy. Today we will talk a little about this algorithm that imitates your calligraphy perfectly.
This device does not seek to falsify the letter of the people, since the people more and more are migrating of the paper towards the screen. However, this is not an impediment for the group of UCL scientists. Achievement puts you thinking of a lot of interesting applications, but you can also go from the practical to the breach of privacy and security.
Perfect imitations
This writing application uses the glyphs, which are graphical representations of a character and at the same time of several characters together, as well as part of a character. When a person writes he resorts to a series of patterns that build those words. Then, the algorithm they created is responsible for detecting these styles, as well as their variations and irregularities, to finally be able to imitate them.
This algorithm that imitates your calligraphy perfectly, basically what it does is add randomness, to ensure that the writing that is replicated does not imitate monotonously and repetitively, that is, can use those variations that people apply when writing .
Also, this program as we were saying is not completely autonomous, with this we mean that it requires the help of an assistant to detect errors in that duplication. The algorithm was named by engineers as My Text in Your Handwriting, as this application goes far beyond what many similar software have achieved, with respect to typographic fonts.
According to Oisin Mac Aodha, one of the scientists after the study, explains that the problem with these fonts is that it is often clear that the text has not been written by hand, which makes it lose touch and personal character Of a handwritten text. What we have developed eliminates that problem.
The algorithm has proved to be of interest to many, so the BBC media world, performed a work on the results interviewing the managers of the same.
In an experiment that was shown to the BBC, to test the wonderful results of the program conducted a test in which several people tried to detect which text was written by the machine and when it was written by the person. Even though people were aware that the algorithm was part of the experiment, 40% of the attempts were mistaken. According to Dr. Tom Haines, this project can be used so that people who can no longer write well can continue to submit manuscript texts, and could be applied for malicious purposes. Even the engineers tested the program with missives from one of the presidents of the United States, Abraham Lincoln, Mexican artist Frida Kahlo and Arthur Conan Doyle and demonstrated those texts could easily pass for theirs.
Similarly, scientists indicate that the technology can also be used to detect fakes. Dr. Gabriel Brostrow commented that we could calculate what percentage of people start their oes from below and not from above, and that type of detailed analysis could reduce dependence on forensics by the forensic services. So this project can be used for the detection of money laundering.