Gaze-driven placement of items for proactive visual exploration

Area of Science:

• Human-Computer Interaction
• Information Visualization
• Digital Signage Systems

Background:

• Digital signage advances enable dynamic item display but lack sophisticated layout schemes.
• Effective item selection within groups on digital displays remains a research challenge.
• Context-aware layouts are crucial for enhancing user experience in visual search tasks.

Purpose of the Study:

• To explore sophisticated item layout schemes for digital signage that respect search context.
• To develop methods for both static and dynamic item placement optimizing visual exploration.
• To infer user preferences and guide dynamic content updates based on eye gaze patterns.

Main Methods:

• Formulated static item placement as an optimization problem with aesthetic constraints.
• Developed dynamic item placement driven by eye-tracking data to infer attractive item progression.
• Utilized matrix factorization on annotated text to extract topics and establish context maps for item replacement.

Main Results:

• Proposed a context-aware planar layout (context map) for dynamic item association and replacement.
• Demonstrated that eye gaze analysis can guide animated layouts towards desired items.
• User studies validated the design criteria for both static and dynamic item placement strategies.

Conclusions:

• The proposed approach offers a framework for sophisticated, context-aware item layouts in digital signage.
• Dynamic, eye-gaze-driven layouts enhance proactive visual exploration and item selection.
• Further research can explore advanced association rules and real-time adaptation for improved user engagement.