gowdaman/gowdaman-emotion-detection

🧠 gowdaman-student-emotion-detection

Model description

This is a fine-tuned version of trpakov/vit-face-expression specifically adapted for analyzing student engagement and understanding during online or in-person educational sessions.

📌 Model Overview

• Model Type: Vision Transformer (ViT)
• Base Model: trpakov/vit-face-expression
• Task: Binary emotion classification for educational sentiment analysis
• Input: Student face image (RGB)
• Output Classes:
  • Understand: The student appears to understand the concept.
  • Not Understand: The student appears to be confused or not following the lesson.

🎯 Use Case

This model is designed for automatic sentiment detection in educational environments, helping instructors evaluate real-time or post-session student engagement by analyzing facial expressions.

It is ideal for:

• Online course engagement monitoring
• Intelligent Learning Management Systems (LMS)
• Post-lecture video analysis for feedback and insights

🧪 Training Details

• Fine-tuned on a custom dataset of student facial expressions captured during course sessions
• Labels were manually annotated as Understand or Not Understand based on visual indicators of comprehension

📥 Input Format

• Image: A single student facial image (preferably frontal and well-lit)
• Size: Automatically resized to match the input size expected by the ViT model

📤 Output

{
  "label": "Understand",
  "confidence": 0.87
}