Adult Facial Fatigue Stress Index

Problem to solve

In workplace health and personal monitoring, fatigue and stress are often judged only by subjective feeling, without a low-friction, recordable reference signal. Facial cues such as under-eye puffiness, dark circles, mouth-corner droop, and glabellar lines can provide observable indicators of state changes. This skill is aimed at smart mirrors, attendance terminals, office health displays, or selfie apps, turning one frontal face image or short video into a structured fatigue/stress score.

How it works

The core capabilities cover frontal face detection, feature quantification, and composite scoring:
- Estimate eye_bag_area for under-eye puffiness or shadow
- Analyze dark_circle_grayscale for darkness around the eyes
- Estimate mouth_corner_drop_deg from the mouth-corner angle
- Score glabellar_frown_lines_score for glabellar lines
- Output fatigue_stress_score (0-100) and levels such as good, mild_fatigue, moderate_stress, and high_stress

A typical run validates the required open-id, accepts a local file or public image/video URL, invokes the skill script, and returns structured results. Inputs should be frontal, well-lit, and free of heavy filters. Supported files include jpg/png images or mp4/avi/mov short videos, typically around a 10MB limit. Historical report queries are fetched from the cloud API and rendered as a table with report name, score, time, and link.

Boundaries and cautions

The score is a personal state reference, not a medical diagnosis or clinical stress assessment. Lighting, makeup, expression, and camera angle can affect a single reading, so continuous trends are more useful. Facial data involves biometric privacy; obtain explicit consent before capture, and apply appropriate access control to images and reports.