Speaker Recognition System

Extract all files in the current MATLAB directory, then type "speakerrecognition" in the MATLAB command window. This will launch a simple and intuitive graphical user interface.

The GUI includes the following functionalities:

1. One-to-One Speaker Recognition - Verification: Select two audio files. The system uses voice feature extraction algorithms (such as MFCCs) to compare vocal characteristics and returns whether they belong to the same speaker. Audio inputs can be loaded from disk or recorded via microphone.

2. Add New Sound to Database: Select an audio file to add to the database. The system requires a unique positive integer ID for speaker association, which will be automatically incremented for organization. Audio can be loaded from disk or recorded using a microphone.

3. One-to-Many Speaker Recognition - Identification: Select an audio file to compare against all voices in the database. The system performs pattern matching using classification algorithms (e.g., GMM or SVM) and returns the recognized speaker ID. The database must contain at least one audio file. Input can be loaded from disk or recorded via microphone.

4. Listen to Sound from File: Load and play audio files directly from disk.

5. Listen to Sound from Microphone: Record and playback audio using the microphone in real-time.

6. Database Information: Display all stored audio files with their corresponding speaker IDs for management and verification.

7. Delete Database: Remove the entire database and all saved audio files to reset the system.

8. Source Code: Access the implementation code for the speaker recognition algorithms and GUI components.

9. README: View the documentation file for system overview and usage instructions.

10. Exit: Close the GUI interface.

This system enables efficient speaker recognition through expandable database management. You can continuously add new audio files to enhance comparison accuracy and speaker identification capabilities.