The objective is to be able to type texts using any device capable of controlling the cursor on the screen. Designed for people with reduced mobility or who have problems using conventional keyboards, it integrates seamlessly with HeadMouse.
A virtual keyboard that predicts the next word, learns from the user, and is capable of saving up to 40% of the strokes necessary to type a text.
VirtualKeyboard uses algorithms that learn from the user's writing style, exponentially improving the rate of success. The system features word prediction functions with algorithms that learn from the user's writing style and exponentially improve the rates of success. The tests carried out by typing literary texts of between 15,000 and 20,000 words demonstrate a saving of up to 40% in the keystrokes necessary to type them, which is equivalent to a reduction of approximately 7,000 words. Use of the keyboard is intuitive and no previous training is required. The virtual keyboard works like an optimized screen application for use by people with motor disability. It supports five dictionaries comprising the most common words in four languages: Spanish, English, French and Italian. Using automatic learning mechanisms, the system picks up new words and can be adapted to the user's writing style, optimizing the prediction of words used in different contexts: emails, technical texts, literary texts, etc.
More than 400,000 downloads of HeadMouse and VirtualKeyboard in 95 countries
Free download
Project developed by the University of Lleida
Related links
