The EasyEyes gaze tracker uses the open-source WebGazer.js package from Papoutsaki et al. (2016). WebGazer and the EasyEyes head tracker both use Google's open-source FaceMesh software (Kartynnik et al., 2019). Other dependencies of each EasyEyes module can be found in its package configuration file in its GitHub repository. The EasyEyes icon is a block print by Cassandra Hussey, used with permission.
As we were wrapping up our development of the EasyEyes Remote Calibrator, we learned of Thomas Pronk's project Eye tracking via a webcam, version 2 that also integrates WebGazer into PsychoJS. The EasyEyes version offers a more stable customized build of WebGazer, including several new features like customized framerate and predicting at the reaction, which is modified to work with other EasyEyes modules and packaged to easily add to any existing experiment pipeline. Another major difference is that the vanilla WebGazer keeps learning from user's interactions even after the calibration phase, which may cause unintentional shifts as participants may look at the fixation while clicking at another position in an experiment. Modified WebGazer in Remote Calibrator allows you to pause learning while keeping prediction.