Privacy user interfaces for eye-tracking Sren Preibusch 20 - PowerPoint PPT Presentation

Privacy user interfaces for eye-tracking Sören Preibusch 20 November 2014 @ W3C Workshop on Privacy and User-Centric Controls

How eye-tracking works Tobii

Time-stamped gaze data plus semantics www.evocinsights.com

Opportunities and applications One-off research tasks Mainstream, continuous  User interface testing (online, physical)  Attention monitoring  Package design  Relevance measurement  Store planning, shelf layout  Ad billing  Drowsiness detection Niche applications  Input modality (for disabled, kiosks)  Various medical diagnosis  Sports performance support  Security, law enforcement

Eye-tracking output data Gaze data Biometrics  Gaze direction and point  User re-identification  Combined with object of interest  Inferred: age, medical condition Eye data Preferences and lifestyle  Eye presence  Eye position and movement  Eyelid closure (blinks)  Pupil size and dilation

Regions of interest Miratech

Regions of interest: by gender Miratech

Potential remedies Semantic APIs  Obstruction and destruction  Meaningful browser APIs  Eye-triggered events: ongazeenter  Meaningful notice to users  :seen pseudo-class, queryable  Meaningful choice and control  Heatmaps: getUserMedia  Policy and regulation Notice and choice  Privacy consequences: explain collected and inferred data  Status indicators (LED, icons)  Hardware and software switches

Thank you. preibusch.de