How adrianvill20/aeye works
AEye is a Windows exam-proctoring application with student and proctor modes. Its documented workflow is: student login and a controlled exam window; per-exam calibration and dual-camera signal capture; comparison with a student-specific behavioral profile; temporal pattern analysis and confidence/severity scoring; then evidence-backed alerts relayed through a central server for human review. Completed exams refine the profile. The tree contains only README.md, so these components and relationships reflect documented capabilities, not inspected implementation or verified code-level wiring.
Main entry points: README.md Signal reading
Components 19
Exam experience
- Student mode
README.md - Student login
README.md - Controlled exam window
README.md
Calibration and sensing
- Exam calibration
README.md - Signal reading
README.md
Behavior analysis
- Student behavior profile
README.md - Behavior learning
README.md - Personalized comparison
README.md - Temporal pattern detection
README.md - Confidence and severity
README.md
Alerts and review
- Flagged moment evidence
README.md - Central server
README.md - Proctor mode
README.md
Other components
- Student
- E-class exam page
- Front camera
- Side camera
- MediaPipe
- Human proctor
Connections 17
- Student → Student mode: uses
- Student mode → Student login: authenticates
- Student mode → Controlled exam window: opens
- Controlled exam window → E-class exam page: displays
- Front camera → Signal reading: provides face feed
- Side camera → Signal reading: provides posture feed
- MediaPipe → Signal reading: extracts face signals
- Exam calibration → Signal reading: supplies zones
- Signal reading → Personalized comparison: provides current signals
- Student behavior profile → Personalized comparison: supplies baseline
- Personalized comparison → Temporal pattern detection: provides behavior
- Temporal pattern detection → Confidence and severity: provides patterns
- Confidence and severity → Flagged moment evidence: scores and flags
- Flagged moment evidence → Central server: sends evidence
- Central server → Proctor mode: relays alerts
- Proctor mode → Human proctor: presents alerts
- Behavior learning → Student behavior profile: updates after exams