This work was done in collaboration with an amazing cross-disciplinary team at Microsoft!

🔑 We show that estimating offsets as users read/verify what they type can significantly enhance typing efficiency and perceived mental workload. Feedback from stakeholders of the ALS community (who are early adopters of the gaze typing technology) has also been very encouraging for us :)

🔍 Miscalibration of eye trackers is a significant barrier to their use and wider adoption. We propose a simple approach to automatically correct for calibration errors during gaze typing, i.e., typing by fixating your gaze on a visual keyboard.

𝗘𝘆𝗲𝗢: 𝗔𝘂𝘁𝗼𝗰𝗮𝗹𝗶𝗯𝗿𝗮𝘁𝗶𝗻𝗴 𝗚𝗮𝘇𝗲 𝗢𝘂𝘁𝗽𝘂𝘁 𝘄𝗶𝘁𝗵 𝗚𝗮𝘇𝗲 𝗜𝗻𝗽𝘂𝘁 𝗳𝗼𝗿 𝗚𝗮𝘇𝗲 𝗧𝘆𝗽𝗶𝗻𝗴
📌 Program: programs.sigchi.org/chi/2025/pro...
🗓 Poster: Monday, April 28 | 6:00 PM - 8:00 PM JST | North 1F
📄 Paper: dl.acm.org/doi/10.1145/...