Livoa LogoLivoa
1. Online Classroom Video Feed


Start of Phase 1: Data Input & Occlusion Handling

2. Facial Complete Capture


Capture individual facial images from video feed

3. Occlusion Check


Is the face occluded?

7. Input: Complete & Facial Classification


Phase 3: Emotion Recognition with CNN

4. Input: Occluded Image


Phase 2: Facial Reconstruction with GAN

5. GAN Processing


Generator Network: Generates complete face

Discriminator Network: Quality check of generated image

Emotion Recognition with CNN
Phase 4: Visualization on the Dashboard


Data Aggregation & Display

Performance Analytics (Confusion Matrix, ROC, F1-Score)

Emotion Analytics (Heat maps, Emotion distribution charts)

Phase 5: Educational Application & Outcomes


Educator Interpretation & Action

Adapt Teaching, Student Support, Enhance Learning

ss

by Snehal

0
0 uses