EduVision: Power your teaching potential 📒

Created for LinceHacks 2024

Authors

• @JocelynVelarde
• @Diego785xd

Features

• Optimized API costs using gpt4-o
• Light and Dark mode enabled
• Available in all devices

Classroom Attention

• Use of computer vision tools to monitor sentiment analysis and attention levels of students
• Fetch data on a dynamic dashboard to view different metrics
• Mantain privacy of students using only tags and not saving biometric data
• Generate reports and reccomendations in relation to the results from lessons
• Identify in which section of the class the students where most engaged, and what was the content being displayed

Generate Content

• Use of generative artificial intelligence to create content for your lessons
• Direct conversation with an assistant to generate the desired content
• Text integration and audios to increase dynamic interaction
• Selection between creating a study plan or lesson content by filling a form
• Visualization of desired topics and pptx creation

Structure

streamlit_app 
├─ home.py
├─ .streamlit
│   └─ secrets.toml
├─ algorithms
│  └─ embeedings.py
│  └─ mongodb_genai.py
│  └─ vectorization_mongo.py
├─ gpt4o
│  └─ image_process.py
│  └─ pptx_genai.py
│  └─ simple_text.py
├─ assets
│  └─ files
│  └─ images
├─ pages
│  └─ Chat.py
│  └─ instructions.py
│  └─ Panel.py
├─ utils
├─ AiManager.py
├─ DataManager.py
├─ index.html
├─ vision_main.py
└─ requirements.txt

Tools

• OpenAI API
• Streamlit
• OpenCV
• MongoDB
• Threading
• Queue
• Deepface
• pandas
• numpy

Deployed with: Streamlit Cloud

Demo

YouTube

License

MIT