This project involves developing a voice detection and recognition system capable of classifying and processing voice commands to control a game environment. The system utilizes MATLAB for audio data acquisition, feature extraction, and classification using machine learning models.
- Data Acquisition: Record voice commands using a microphone and store them in a predefined directory structure.
- Feature Extraction: Preprocess audio data and extract relevant features for classification.
- Machine Learning: Train models such as K-Nearest Neighbors (KNN) and Neural Networks to classify voice commands.
- Real-Time Testing: Test the models in real-time to control a game using voice commands.
- Interactive Game: Control elements in the game (e.g., move a cursor) using recognized voice commands.
- MATLAB R2023 or later.
- Audio recordings of commands stored in
sons_audio. - Microphone for real-time testing.
- Clone the repository:
git clone https://github.com/Abib-web/Voice-Detection-and-Real-Time-Control-Project.git cd Voice-Detection-and-Real-Time-Control-Project - Open MATLAB and navigate to the project directory.
- Ensure all required dependencies are installed in MATLAB.
Run the GEI1090_Prog_Acquisition.m file to record voice commands.
The commands will be stored in sons_audio with appropriate labels.
Execute the GEI1090_Prog_Traitement.m file to preprocess the data, extract features, and train machine learning models.
Run the GEI1090_Prog_Traitement.m script in real-time testing mode to control the game using voice commands.
Use voice commands like "Haut," "Bas," "Gauche," and "Droite" to control the cursor in the game environment.
Below is an example of the game environment controlled by voice commands:
GEI1090_Prog_Acquisition.m: MATLAB script for data acquisition.GEI1090_Prog_Traitement.m: MATLAB script for processing and training models.sons_audio/: Directory for storing audio recordings.
- Oumar Kone
This project is licensed under the MIT License. See the LICENSE file for more details.