TurkerGaze: Crowdsourcing Saliency with Webcam based Eye Tracking

Abstract

Traditional eye tracking requires specialized hardware, which means collecting gaze data from many observers is expensive, tedious and slow. Therefore, existing saliency prediction datasets are order-of-magnitudes smaller than typical datasets for other vision recognition tasks. The small size of these datasets limits the potential for training data intensive algorithms, and causes overfitting in benchmark evaluation. To address this deficiency, this paper introduces a webcam-based gaze tracking system that supports large-scale, crowdsourced eye tracking deployed on Amazon Mechanical Turk (AMTurk). By a combination of careful algorithm and gaming protocol design, our system obtains eye tracking data for saliency prediction comparable to data gathered in a traditional lab setting, with relatively lower cost and less effort on the part of the researchers. Using this tool, we build a saliency dataset for a large number of natural images.

Paper

• P. Xu, K. A. Ehinger, Y. Zhang, A. Finkelstein, S. R. Kulkarni, and J. Xiao.
  TurkerGaze: Crowdsourcing Saliency with Webcam based Eye Tracking
  arXiv:1504.06755v1 [cs.CV] 25 Apr 2015.

Video (Download raw video here)

Other Materials

• Supplementary Materials
• Poster for VSS

Source Code and Data

• Data: Refer to the Saliency Prediction in the LSUN challenge http://lsun.cs.princeton.edu.
• Source Code: Available at GitHub (Old version. We plan to update a new/better version in Spring 2016.)

Acknowledgement

This work is supported by gift funds from Google, MERL, and Project X to JX.