TinyHAR, a 46k-parameter time-series model, reaches Macro F1 0.960 on 5-way gaze gesture recognition and 0.997 on user identification using ARKit data in a 4-person pilot.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.HC 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
TinyGaze: Lightweight Gaze-Gesture Recognition on Commodity Mobile Devices
TinyHAR, a 46k-parameter time-series model, reaches Macro F1 0.960 on 5-way gaze gesture recognition and 0.997 on user identification using ARKit data in a 4-person pilot.