GlamAR's AI Facial Skin Analysis delivers real-time skin diagnostics using camera-based input. To ensure maximum accuracy, performance, and user trust, the following device, camera, lighting, and network conditions must be met.
Platform | Environment |
---|
Android SDK | Android 7.0+ (API 24+), recommended ≥ Android 10 |
iOS SDK | iOS 13+ (native only, Web SDK not recommended) |
React Native | Android & iOS (uses native camera modules) |
Flutter | Android & iOS via InAppWebView |
Web SDK | ⚠️ Supported for POCs only — not recommended for production due to browser camera quality limitations |
Smart Mirrors | Android-powered with external USB camera support |
❗ Web SDK is not recommended for live skin scans due to variable image quality. Use native or mirror-based integrations.
Device Compatibility
Tablet Support
OS | Support Level | Notes |
---|
Android | ✅ Full Support | Works with native SDK or Chrome |
iPadOS | ⚠️ Limited Support | Safari camera APIs |
Device Model | Details |
---|
iPad Pro (2021+) | 12MP front camera + Apple Neural Engine |
Samsung Galaxy Tab S8/S9 | Excellent selfie cam + strong processors |
Lenovo P12 Pro / Yoga Tab | Mid-range Android options with decent optics |
Mount device in landscape with centered face framing for best detection.
Smart Mirror Support
Component | Recommendation |
---|
Camera | Logitech Brio 4K / Intel RealSense D415 |
CPU | Intel i5/i7 10th Gen+, or macOS M1/M2 chip |
RAM | Minimum 8 GB |
GPU | Not required unless running on-device ML |
Display | ≥15", Full HD, touch-enabled preferred |
🛠️ Our Android SDK can run on most kiosk builds. Reach out for mirror integration guides.
Enviroment & Lighting
Environment Guidelines
Setting | Recommended Practice |
---|
Indoor Use Only | Avoid direct sunlight, window glare |
Subject Distance | 12–18 inches (30–45 cm) from the camera |
Hair & Face | Tie hair back, remove makeup, no glasses |
Face Framing | Head upright, eyes open, mouth relaxed |
🎯 These controls ensure model alignment and reduce false positives.
Lighting Conditions
Parameter | Recommended Setup |
---|
Intensity | 500–1000 lux ambient lighting |
Type | White LED with 4000K–5000K color temperature |
Avoid | Backlighting, overhead glare, or tinted light |
Suggested Setup | Ring light or diffused LED panel at ~45° angle in front of subject |
🎯 Consistent lighting improves analysis of redness, pigmentation, and under-eye features.
Camera
Camera Requirements
Parameter | Specification |
---|
Resolution | Minimum 8MP (recommended ≥12MP) |
Focus | Auto-focus required; avoid fixed-focus webcams |
Frame Rate | Minimum 30 FPS |
Color Reproduction | Accurate white balance; no beauty filters |
Sensor Type | CMOS or higher-end imaging sensors (preferred) |
🔍 Why this matters: Skin conditions like pores, acne, and wrinkles require fine detail. Low-res or beautified images reduce detection accuracy significantly.
Camera Permissions
Platform | Permissions Needed |
---|
Android SDK | android.permission.CAMERA |
iOS SDK | NSCameraUsageDescription in Info.plist |
React Native | Use react-native-permissions |
Flutter | Use permission_handler plugin |
Web SDK | Standard browser-based camera access prompt |
📱 Always ask permission before triggering camera. iOS requires camera usage description in app metadata.
Network Requirements
Metric | Minimum | Recommended |
---|
Upload Speed | ≥2 Mbps | ≥5 Mbps |
Latency | < 300 ms | < 150 ms |
Offline Support | Optional* (via edge caching) | Contact us |
Metric | Typical Value |
---|
Capture → Result | ~1.5s–3s on mobile / ~800ms–1.5s via API |
Image Format | JPEG or PNG (base64 encoded for API usage) |
Annotated Output | 1 base64 image per condition + original |
*Offline/local scan support is available for enterprise mirrors via local inference or edge caching.