SmartEye: An Accurate Infrared Eye Tracking System for Smartphones
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
The capability to estimate where a user is looking on a screen is known as gaze estimation or eye tracking. It has been used in medical applications including assessment of mood and learning disorders, and brain injury diagnosis. If accurate eye tracking could be integrated into commodity smartphones these diagnostics could be broadly deployed at very low cost. The highest accuracy and most robust eye tracking methods employ infrared cameras and illumination which are not yet available on all standard smartphones. In this paper, we present an accurate infrared eye tracking system on a smartphone, named SmartEye, on an industrial prototype phone equipped with an infrared camera and illumination. The system is accurate in the presence of head pose variation and device movements in the user's hands, and requires only a one-time calibration routine to measure specific parameters of the user's eye. Our system achieves a gaze estimation bias of 0.57 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">°</sup> at a 20cm distance from the user, 5 times better than state-of-the art mobile device eye-tracking systems that do not use infrared illumination. Our system also allows for free head movements at distances between 20-40cm with a moderate increase in average gaze bias (to ~1 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">°</sup> ), and can operate at 12fps. This enhanced accuracy and increased mobility can expand significantly the range of eye-tracking applications that can be supported by smartphones.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it