Skip to content

Latest commit

 

History

History
47 lines (32 loc) · 1.83 KB

README.md

File metadata and controls

47 lines (32 loc) · 1.83 KB
Raw

TurkerGaze

TurkerGaze: Crowdsourcing Saliency with Webcam based Eye Tracking

Instructions

> ./scripts/install.sh
> npm install
> npm start

Created by Pingmei Xu, Jianxiong Xiao at Princeton Vision Group.

Introduction

TurkerGaze is a webcam-based eye tracking game for collecting large-scale eye tracking data via crowdourcing. It is implemented in javascript and the details were described in an arXiv tech report.

system overview

Citing

If you find TurkerGaze useful in your research, please consider citing:

@article{xu15arXiv,
    Author = {Pingmei Xu, Krista A Ehinger, Yinda Zhang, Adam Finkelstein, Sanjeev R. Kulkarni, Jianxiong Xiao},
    Title = {Rich feature hierarchies for accurate object detection and semantic segmentation},
    Booktitle = {arXiv:1504.06755},
    Year = {2015}
}

Usage

  1. See a demo 0. Setup a local web server, download the folder, open 'pugazetrackr.html' to run the eye tracking task. Save the result data to a local file and visualize the result by 'visualizer.html'. 0. You can also try it here: eye tracking task visualization

  2. User your own images 0. Create a .json object with two fields: 'gaze' and 'memory' like './demo/imglist.json'. 'gaze' contains the images that you want to present for free-viewing, and 'memory' contains images for the memory test. 0. Pass the path of this .json file by url parameter 'imglist'. For example, http://isun.cs.princeton.edu/TurkerGaze/pugazetrackr.html?imglist=your_imglist_url 0. Run the task!

License

TurkerGaze is released under the MIT License (refer to the LICENSE file for details).