• The script includes error handling mechanisms to handle cases where speech recognition fails to understand the user's command or if there are issues with the network or ESP8266 communication.
• In case of errors or unrecognized commands, the script prompts the user to repeat the command or provides assistance in understanding the available commands.

6. Continuous Operation

• The script runs in a continuous loop, constantly listening for voice commands from the user.
• It remains active until terminated by the user, allowing for seamless interaction and control of the home automation system.

In summary, the code along with the website enables voice-controlled home automation by leveraging speech recognition to interpret user commands, integrating with an ESP8266 module to control lights remotely, and providing audible feedback to the user regarding the status of their commands. The system operates continuously, allowing for intuitive and hands-free control of lighting systems within the home.

Features

Machine Learning Models

• Pre-trained Models: Hosts a collection of pre-trained models for various AI tasks, ready to be used out of the box.
• Custom Model Training: Provides tools and utilities for training custom machine learning models on user-provided data.
• Text Generation: Generate coherent text based on prompts using state-of-the-art language models.

Integration with ESP8266 for Remote Light Control

• Remote Light Control: LumosAI_ integrates with an ESP8266 model to allow users to remotely control lights.
• Toggle Control: Users can toggle lights on and off remotely through the project website.
• Voice Control: Voice commands are processed by the AI model to control lights via the ESP8266 module.

Getting Started

To get started with LumosAI_, follow these steps:

Pre-trained AI setup

1. Clone the Repository:

   git clone https://github.com/R3tr0gh057/LumosAI_.git
   

2. Install Dependencies:

   cd LumosAI_/AI
   pip install -r requirements.txt
   

3. Setup:

   • Connect your ESP8266 module to your local network and note down its IP address.
   • Update the URLs in the script to match the IP address and endpoints of your ESP8266 module.
   • You can change the hotword by changing the value of the hotword variable.

4. Usage:

   • Run the script and wait for it to start listening.
   • Say you hotword to activate the AI.
   • Proceed to use the commands like "turn on the bedroom lights" or turn on the hall room lights".
   • You can customize the script to add more voice commands or integrate additional functionality based on your requirements.

Custom Model Training (Beta)

1. Install Dependencies:

   cd LumosAI_/hotword
   pip install -r requirements.txt
   

2. Usage:

   python PreparingData.py
   

   This will prompt you to record yourself saying the hotword 100 times as default, you can change the amount of samples by adjusting the parameters.

   python PreProcessingData.py
   

   This will create the audio data upon which the ai will be trained.

   python training.py
   

   This will use the audio data and create a trained model which can be used to determine if the wake word is present in a recorded voide note.

   python RunParallely.py
   

   This will run a program that uses the trained data to compare the newly recorded audio data's spectrogram and compares it with the trained data points. If the data points match, it will return true and asks the user for the query. To just test the training data, use the following:

   python prediction.py
   

   This will tell you if the custom hotword is present in the newly recorded data.

Ngrok Setup

Setting up your ngrok port forwarding service is essential since direct direct http communication from an https server is not possible. To overcome this, we can use a port forwarding service to forward our http port to an https domain.

1. Setup:

   Open an administrator cmd and run the following commands:

   choco install ngrok
   ngrok config add-authtoken <TOKEN>
   ngrok http --<yourfreestaticport> http://ip:port/
   

2. Note:

   • Make an ngrok account and replace with your auth token
   • Replace with your free static port offered by ngrok
   • Once its set up, make sure you're connected to the same WiFi as the one connected to the esp8266

ESP8266

You must make a few changes to the esp8266 code in order for it to connect to your WiFi and enable server communication. Here are the changes that you have to make:

1. WiFi:

   • Change the and to your respective wifi name and password
   • Make sure your WiFi is set to 2.4 gHz

2. Server:

   • You can change the endpoints of your server depending on your needs by making changes to server.on(/, HTTP_GET, ), change to your desired endpoint.

WEBPAGE

You need to change the javascript to change the endpoints to establish communication with the esp8266 module. You should make sure you have good internet connection for quick responses. Here are the necessary changes:

• Change the fetch requests of "ai.js" to the the following format: , for example: 'https://bulldog-promoted-accurately.ngrok-free.app/power-saver1'
• Change the firebase config variable of "login.js" and "script.js" with your firebase config code which will be available on your firebase profile

License

This project is licensed under the MIT License. Feel free to use, modify, and distribute the code according to the terms of the license.

Contact

For inquiries or feedback, please contact the project maintainer:

• Maintainer: Joe Sanjo Cijo
• Email: joesanjo10@gmail.com
• GitHub: R3tr0gh057