Creating a smooth and efficient way to enter text without a traditional keyboard continues to be a tough hurdle in the world of extended reality (XR) headsets. Researchers have now put together an extensive collection of existing text input techniques, analyzing their strengths and weaknesses. By offering this catalog to the public, they hope to kickstart further innovations and improvements in this area.
Maximilian Di Luca, an Associate Professor at the University of Birmingham, manages the VR Lab there and has previously worked with Meta to advance hand input and haptics in VR. His latest work, a joint effort with industry experts, was honored at the ACM SIGCHI 2025 awards for its groundbreaking input framework for XR devices.
As virtual and augmented reality experiences become more advanced, one significant hurdle remains: effective text input. Whether it’s composing emails in a virtual office or logging into a digital world, the need for quick and accurate text entry is essential for XR applications.
Our team at the VR Lab, along with collaborators from several esteemed institutions and Google, tackled this challenge by developing the XR TEXT Trove. This project documents over 170 text input techniques specific to XR, providing a detailed evaluation of each method’s pros and cons. The TEXT Trove stands as an instrumental resource for identifying the most effective text entry approaches, both in academic and commercial settings.
The techniques in our catalog are classified under 32 categories. These include 13 interaction attributes like input devices and body part usage, as well as 14 performance metrics, including speed and error rates. This comprehensive breakdown helps paint a clear picture of the current landscape of text input methods in XR.
One crucial insight from our research is that text input speed is closely tied to the number of active input elements. Only when multiple fingers are used can we achieve speeds comparable to traditional typing on keyboards. Each additional finger brings a noticeable increase in words per minute.
Visual data from our research underscores the fact that varied input devices and techniques yield different results. Our plots illustrate the range of text input speeds possible with different configurations, highlighting the potential for increased efficiency through multi-finger use.
Haptic feedback and the use of external surfaces emerge as key factors in enhancing typing performance. Typing on solid surfaces, rather than in mid-air, offers a more comfortable experience and reduces muscle strain, which can lead to issues like “Gorilla Arm Syndrome.”
Interestingly, the classic keyboard layout remains unmatched for sheer typing speed, likely due to its familiar structure and learning curve. But the search for a faster method continues. Techniques like machine learning and AI could potentially shorten finger travel distances, akin to the breakthrough of swipe typing on smartphones.
The XR TEXT Trove marks a significant stride toward more effective text input in the realms of virtual and augmented reality. Our goal was to provide a valuable tool for researchers and developers, enabling the next generation of user-friendly text solutions in immersive environments.
We’re excited to share our full paper at the upcoming ACM CHI conference in Yokohama, Japan. Alongside this project, some team members have also contributed to the Locomotion Vault, a similar initiative focused on cataloging VR movement techniques, which further aids researchers and designers in optimizing XR environments.
