https://github.com/qonus/ai-teacher

Project suggests using AI Teachers for efficient education!!
https://github.com/qonus/ai-teacher

Last synced: about 1 month ago
JSON representation

Project suggests using AI Teachers for efficient education!!

• Host: GitHub
• URL: https://github.com/qonus/ai-teacher
• Owner: Qonus
• Created: 2024-06-19T17:35:29.000Z (8 months ago)
• Default Branch: main
• Last Pushed: 2024-07-12T22:15:36.000Z (7 months ago)
• Last Synced: 2024-11-08T03:05:56.868Z (3 months ago)
• Language: Python
• Size: 11 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# AITeacher

## Description:

 - This is project developed during Summer Research Program in ISSAI NU (Nazarbaev University in Astana) by Qonus✌️😎, riz3e and himoji.

 - The Idea of the project is to teach young generation with interactive AI kazakh teachers.

 - We plan on improving the original prototype from ISSAI by implementing command system into our AI teacher. Theoretically it will give teacher more freedom and immerse the student into realm of education. Teacher will be able to show relevant media (Images. maybe videos too?), perform animations, change voice with speeking rate and show emotions with AI generated face animations on command. Chat bot will ocasionally use those commands.

## Instructions:

 - Clone or download the project from this repository.

 - Install dependencies listed in requirements.txt.

 - Enter `python main.py` in project directory terminal to run the program (I'm so sorry for people who got an error message after that, good luck...!).

 - Choose preferred input device by typing index shown in terminal.

 - Hold [SPACE] key to record yourself and ask AI a question.

 - Enjoy! Additionally, you can add new AI behavior inside `teacher_personalities/` directory and reference it in 7th line of `main.py`, here: `PERSONALITY_FILE = r"AITeacher/teacher_personalities/Kakashi.txt"`.

## Currently used api list:

 - (stt.py) faster-whisper for speech to text.

 - (gpt.py) chat gpt 4 as chat bot, the big brain behind our AI.

 - (tts.py) edge-tts for text to speech.

 - (tti_generation.py) hugging face's stable diffusion models for text to image.

 - (tti.py) unsplash api for database of stock images.

## Note:

 - There are other scripts labeled "old". They are old solutions I left in project. During development process I had to get rid of them because the solutions were whether too slow just like whisper, or not very accurate/real like in the case of pyttsx3, llama and SpeechRecognition.

 - server.py will be relevant for visual representation of AI soon enough. We will use it to send to and possibly receive messages from Unreal Engine (3D game engine with realistically real graphics, in fact you can't say Unreal Engine without saying "real").