Privacy-first Android app
Personal pain tracking with confidence-aware insights, calm guidance, and local-first analysis.
Know how reliable each capture is.
See how visible change develops over time.
Guidance helps improve pose and lighting.
Daily overview
Personal baseline comes first.
Population similarity stays secondary and confidence-aware.
Which label best matches this face?
Baseline first
PainTrace is built around the user’s own baseline. The app highlights visible deviation, capture quality, trend direction, and confidence context, while keeping population similarity secondary and cautious.
Current captures are compared against previous local baseline information, so change is read in relation to the same person over time.
Lighting, blur, face size, frontal pose, occlusion, and one-face checks help users understand whether a capture is strong enough to interpret.
The app supports gentle tracking of visible change, rather than making isolated certainty claims from one image.
Privacy
Raw photos and raw audio are optional. PainTrace is designed for careful data handling, local-first processing, export controls, and deletion controls.
The first version is designed without a cloud backend, reducing unnecessary data movement.
The app is positioned without advertising or behavioural analytics in the initial design.
Facial feature review is used for change monitoring, not for identifying people.
Research
PainTrace’s direction is shaped by privacy-first, cautious signal interpretation and structured non-diagnostic review.
Classifying Vocal Expressions of Physical vs. Emotional Pain Using Offline AI: Concept and Methodology
Jakub Tencl, published May 4, 2025 on Zenodo. The work provides methodological context for privacy-conscious pain-related signal analysis.
DOI: 10.5281/zenodo.15337886