iCare 日照機器人

Overview

iCare: A System for Conversing with the Elderly to Alleviate Dementia

1. Utilizes Deepface and OpenCV for emotion recognition, and Azure Speech to Text to convert speech into text.
2. The recognized emotions and converted text are used as inputs to generate responses through the OpenAI API (GPT).
3. The generated response is then passed to MuseTalk to create a virtual avatar.

待開發功能

吃藥回診提醒:alert
語音 style:style

一律推到 branch 確定再 push 回 main

Technologies Used

1. Azure Speech to text
2. OpenAI API
3. MuseTalk
4. Deepface & OpenCV

Installation Guide

1. Download MuseTalk and place it in your folder.
2. Place both in the same folder with the following structure:

   icare/
   ├── MuseTalk/
   ├── main.py
   ├── requirements.txt
   ├── templates/
   └── static/
   

3. Create a new environment:

   conda create --name myenv python=3.8
   

4. Switch to the environment and run the following command:

   pip install -r requirements.txt
   

5. Get API keys of Azure, OpenAI. Then put them in .env file.
6. Run the program:

   python main.py
   

Acknowledgment

Special thanks to MuseTalk for serving as the model for virtual avatar output.