Percepta
Vision reimagined — a socioemotional guide that reads facial expressions in real time, entirely on your iPhone.
Why Percepta
For many people — with low vision, on the autism spectrum, or with social-emotional agnosia — facial expressions are hard or impossible to read. Percepta points the front camera at a conversation partner and identifies the expression as it happens: spoken aloud for VoiceOver users, shown on screen, and felt as a gentle haptic tap.
Features
Live expression reading
Happy, Sad, Surprised, Angry, Neutral — with a confidence meter and per-feature detail for smile, eyes, and brows.
Feel it, don't watch it
A gentle haptic tap plays when the expression changes, with adjustable strength in Settings.
Session history
See which expressions appeared in a conversation and for how long — stored only on your device.
Private by design
All perception runs on-device with Apple's Vision framework. No video is recorded, stored, or uploaded — ever.
Built for accessibility
Full VoiceOver support, Reduce Motion respected, and live readings announced — not just shown.
Spatial guide — coming soon
Real-time obstacle and environment awareness from the iPhone camera, delivered as audio and haptic cues. In development.
How it works
- Sign in with email, Google, or Apple.
- Allow camera access — required for the core feature.
- Point the front camera at a face. Percepta aligns, reads, and reports the expression in real time.
- Turn-toward-the-camera guidance appears when a face is too far off-angle for a reliable reading.
Honest limits
Percepta detects facial movements — a smile, raised brows — and labels the expression they most resemble. Facial expression is not the same as inner feeling, detection is approximate, and it works best with a single, well-lit face. Percepta is an assistive aid, not a medical device.