A Gaze-depth Estimation Technique with an Implicit and Continuous Data Acquisition for OST-HMDs

Y. Lee , T. Piumsomboon , B. Ens , G. A. Lee , A. Dey , M. Billinghurst

ICAT-EGVE (Posters and Demos)

Optical see-through headsets need to know how far away a user is looking, but measuring gaze depth usually requires explicit calibration. This paper proposes an implicit, continuous data-acquisition method that constantly gathers gaze data — including inter-pupillary distance and the gaze distance to real and virtual targets for each eye — to train a machine-learning model for 3D gaze-depth estimation. The approach aims to make depth-aware gaze interaction work smoothly in everyday AR use.

@inproceedings{lee2017,
  author    = {Y. Lee and T. Piumsomboon and B. Ens and G. A. Lee and A. Dey and M. Billinghurst},
  title     = {A Gaze-depth Estimation Technique with an Implicit and Continuous Data Acquisition for OST-HMDs},
  booktitle = {ICAT-EGVE (Posters and Demos)},
  pages     = {1-2},
  doi       = {10.2312/egve.20171364},
  year      = {2017}
}