EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·Aug 20·5 min read
Seeing beyond BMI: Estimating cardiometabolic risk with smartphone imagery
TL;DR
PhotoScan, a deep learning framework developed by Google Research, estimates body composition from smartphone photos, achieving insulin resistance predictions comparable to DXA scans.
✦ Why It Matters
Consider integrating PhotoScan into health assessments to enhance early detection of metabolic risks.
Key Takeaways
- Implement PhotoScan for non-invasive body composition analysis.
- Utilize smartphone imagery to enhance metabolic health assessments.
- Combine demographic data with PhotoScan for better insulin resistance predictions.
- Explore multi-modal data integration for comprehensive health monitoring.
How It Works
PhotoScan uses deep learning to analyze 2D smartphone images, extracting geometric body information to estimate three-dimensional body composition metrics like body fat percentage, Android-to-Gynoid fat ratio, and Visceral-to-Subcutaneous fat area ratio.