Touchscreens are notoriously difficult to type on. Since we are not-able to feel the keys, we believe the sense of sight to maneuver our fingers to the proper places and check for errors, a com-bination of tasks that’s difficult to accomplish simultaneously. to actually understand how people type on touchscreens, researchers at Aalto University & Finnish Center for AI (FCAI) have created the 1st AI model that predicts how people move their eyes and fingers while typing.
The AI model can simulate how a person’s user would type any sentence on any keyboard design. It makes errors, detects them—though not always immediately—and corrects them considerably like humans would. The simulation also predicts how people adapt to alternating circumstances, like how their writing-style changes once they start using’ a new auto-correction system or keyboard design.
“Previously, touchscreen typing has been understood mainly from the mindset of how our fingers move. AI-based methods have helped shed new light on these movements: What we’ve discovered is that the importance of deciding when and where to seem . Now, we-can-make far better predictions on how people type on their phones or tablets,” says Dr. Jussi Jokinen, who led the work.
The study, to be presented at ACM CHI on 12 May, lays the groundwork for developing, as an example , better and even personalized text entry solutions.
“Now that we’ve a sensible simulation of how humans type on touchscreens, it should be more easier to optimize keyboard designs for better typing—meaning fewer errors, faster typing, and, most significantly on behalf of me , less frustration,” Jokinen explains.
In addition to predicting how a generic person would type, the model is also-ready to account for various sorts of users, like those with motor impairments, and might be used-to develop typing aids or interfaces designed with these groups in mind. For those facing no particular challenges, it can deduce from personal writing styles—by noting, as an example , the mistakes that repeatedly occur in texts and emails—what quite a keyboard, or auto-correction system, would best serve a user.
The novel approach builds on the group’s earlier empirical-research , which provided the idea for a cognitive model of how humans type. The researchers then produced the generative model capable of typing independently. The work was done as a part of a bigger project on Interactive AI at the Finnish Center for AI .
The results are underpinned by a classic machine learning method, reinforcement learning, that the researchers extended to simulate people. Reinforcement learning is generally used-to teach robots to unravel tasks by trial and error; the team found a new-way to use this method to get behavior that closely matches that of humans—mistakes, corrections & all.
“We gave the model same -to-same abilities and bounds that we, as humans, have. once we asked it to type efficiently, it found out the way to best use these abilities. the final result’s very almost like how humans type, without having to-tech the model with human data,” Jokinen says.
Comparison to data of human typing confirmed that the model’s predictions were accurate. In-the future, the team hopes to simulate slow and fast typing techniques to, for instance , design useful learning modules for people that want to enhance their typing.
The paper, “Touchscreen Typing As Optimal Supervisory Control,” was presented 12 May 2021 at the ACM CHI conference.
The findings were reported on Aalto University