AI-Teacher

Inspiration

The inspiration behind AI-Teacher stems from the struggles students face during high-pressure examination periods. Lengthy PDFs and PPTs are hard to process within tight deadlines, and external resources like YouTube often fail to align with syllabus requirements. AI-Teacher was designed to bridge this gap, making education personalized, engaging, and efficient.

What it does

AI-Teacher offers a revolutionary learning experience through:

  • 3D Virtual AI Teacher: An interactive avatar capable of real-time lip-syncing, AI teaching content from uploaded PDFs or PPTs with synchronized audio (text-to-speech) and text on a virtual board.
  • Doubt-Solving Feature: Provides instant answers to student queries using GenAI.
  • Activity Tracker: Visualizes learning streaks and progress with a heatmap graph, motivating consistent study habits.

How we built it

Backend: AI-driven NLP models (Google Gemini) to analyze uploaded content and convert it into structured lessons.

  • Frontend: A 3D avatar using Ready Player Me and real-time lip-syncing with TTS for interactive teaching.
  • Additional Features: Ask me doubt feature, which solves students doubt on the taught lesson on the same time.

Challenges we ran into

  • Implementing real-time lip-syncing for the 3D avatar was complex and required fine-tuning synchronization.
  • Developing long AI generated video requires high computational power
  • Ensuring seamless analysis of diverse document formats for effective teaching.

Accomplishments that we're proud of

  • Successfully creating a lifelike 3D virtual teacher capable of interactive, syllabus-aligned lessons.

  • Providing a complete solution that integrates teaching, doubt-solving, and progress tracking seamlessly.

  • New way of learning for New Generations. (i.e AI BASED LEARNING)

What we learned

  • Advanced techniques in natural language processing for extracting key information from educational content.

  • Effective integration of TTS and 3D avatar technologies for interactive learning.

  • Large Language Model

What's next for AI-Teacher

  • Expanding subject coverage by integrating more advanced models tailored to different academic disciplines.
  • Enhancing the 3D avatar with multilingual capabilities to support diverse student needs.
  • To also scan and teach on the handwritten notes provided by student using OCR technology.

Built With

Share this project:

Updates