Objective

To build an AI-based system that analyzes students’ facial expressions during online classes using their webcam, to assess their emotional engagement, attention span, and identify potential learning difficulties in real-time.

Key Features

Student Side

• Consent-based webcam access during online sessions.

• Real-time facial emotion recognition (happy, confused, bored, focused, etc.).

• Emotion log timeline available to review attention levels.

• Privacy-focused: Data is locally processed or anonymized.

Instructor Panel

• Real-time emotion dashboard for all active students.

• Alerts if a student shows confusion, boredom, or disengagement for extended time.

• Emotion trends visualized per student (charts/graphs).

• Export session-wise emotion summary.

Admin Panel

• Manage classes, students, and instructors.

• System logs for facial detection and error tracking.

• Configuration settings for detection sensitivity, model updates, etc.

• Data compliance tools (anonymization, opt-out, logs).

Tech Stack

Workflow (Step-by-Step)

1. User Onboarding

• Students and instructors sign up and log in securely.

• Students join live classes through the web app (Zoom/Meet integration optional).

2. Live Class with Emotion Tracking

• With consent, the student’s webcam is accessed via browser.

• Emotion detection model runs in the background:

  • Facial expressions are captured and analyzed in real-time.

  • Detected emotions are mapped to a timeline.

• Processing is done on-device or sent securely to backend for analysis.

3. Instructor Dashboard

• Instructors see a real-time heatmap of student emotions.

• System notifies when multiple students show negative emotions (e.g., confusion).

• Emotion charts after the session show patterns:

  • Time spent attentive

  • Emotional spikes

  • Drop in engagement

4. Admin Insights

• Access to anonymized analytics across all sessions.

• Download reports of emotional engagement by class/session.

• Manage model configurations and update classifiers if needed.

Privacy & Ethics

• Emotion tracking is opt-in, and no video is stored.

• Only emotion labels + timestamps are stored for analysis.

• GDPR and educational data compliance enforced.