Guided Frame 2.0
A research-driven accessibility feature empowering blind and low vision users to capture and share visual moments. Featured in the Google Superbowl 2024 commercial.
Duration
8 weeks
Industry
Mobile Accessibility
Methods
Field Study, Usability Testing
Participants
Blind & Low Vision Users
Research Background
Guided Frame 2.0/3.0 was designed to address the unique needs of blind and low vision users, leveraging Pixel's advanced sensors, camera, and AI. Foundational and evaluative research revealed opportunities to empower users in capturing their surroundings and sharing visual stories.
Challenge: Making photo capture and framing accessible for users with little or no sight, while ensuring independence and confidence.
Key Research Questions
- How do blind and low vision users approach photo capture and framing?
- What guidance and feedback mechanisms are most effective for independent use?
- How can AI and sensors be leveraged to improve accessibility and user experience?
- What are the barriers to adoption and satisfaction for this user group?
Research Methodology
Field Study & Usability Testing
- • Contextual inquiry in real-world settings
- • Think-aloud protocol and dual photo rating
- • Accessibility testing with blind and low vision participants
- • Feedback on guidance cues, descriptions, and outdoor photo support
AI & Sensor Integration
- • Camera/face tilt cues and lighting guidance
- • Dynamic descriptions and improved person/object detection
- • Integration of Gemini for richer contextual feedback
- • Cross-functional collaboration and participant safety
Key Research Insights
Empowering Independence
Guided Frame enabled users to capture photos independently, increasing confidence and willingness to share visual stories.
Effective Guidance Cues
Camera/face tilt cues and dynamic descriptions improved photo quality and user satisfaction.
AI-Driven Accessibility
Gemini integration provided richer, contextual feedback, especially for outdoor photos and complex scenes.
Research-Driven Recommendations
Accessible Guidance & Feedback
- • Refine camera/face tilt cues for clarity
- • Expand dynamic descriptions for more scenarios
- • Prioritize feedback for outdoor and group photos
- • Continue usability testing with diverse participants
AI & Sensor Innovation
- • Enhance Gemini integration for richer feedback
- • Improve person/object detection accuracy
- • Explore new use cases for AI-driven accessibility
- • Foster cross-functional collaboration
Research Impact
Research findings from field studies and usability testing directly informed product changes across guidance, feedback, and AI descriptions. These updates improved independent use, photo quality, and overall satisfaction for blind and low-vision participants and unlocked new product opportunities for a broader audience.
Increase in independent use
Sixty percent increase in independent use after design changes
+45%
Average photo quality improvement
Forty five percent average improvement in photo quality metrics
92%
Feature satisfaction (survey)
Ninety two percent feature satisfaction reported in post-launch surveys
+30%
Increase in user confidence
Thirty percent increase in user confidence measured during follow-up testing
Notes: metrics are aggregated from mixed-methods studies (field observations, lab testing, and surveys) and reflect observed improvements after implementing recommended design and ML changes.
More Research Projects
Pixel Accessibility Product | Google Superbowl 2024 Commercial.
Research that improved independent photo capture for blind and low-vision users, informing product and ML improvements.
Foundational research for sustainable tourism and brand strategy.
Mixed-methods research to shape a sustainable tourism brand and stakeholder strategy for Kosovo.
Assistive magnification app research for Android.
Accessibility heuristics, concept testing, and optimization that increased adoption and usability for low-vision users.