README.md

metadata

title: Voice Access Control System
emoji: 🎤
colorFrom: indigo
colorTo: purple
sdk: gradio
sdk_version: 5.12.0
app_file: app.py
pinned: false

Voice Access Control System

This is a deep learning-based voice access control system that can verify whether a person should be granted access based on their voice recording.

Description

The system uses a convolutional neural network to analyze mel spectrograms of voice recordings and determine if the speaker is authorized. It processes audio input through several steps:

1. Audio preprocessing (resampling, normalization)
2. Mel spectrogram generation
3. Deep learning model analysis
4. Access decision with confidence score

Usage

1. Click the audio input button or drag and drop an audio file
2. Wait for the system to process the recording
3. View the access result and confidence score

Technical Details

• Model: Custom CNN architecture (VoiceAccessNet)
• Input: Audio files (WAV, MP3)
• Audio processing: 16kHz sample rate, mel spectrogram features
• Output: Binary classification (Access Granted/Denied) with confidence score

References

• Model training code and dataset details: [Link to your repository]
• Based on PyTorch and torchaudio
• Deployed using Gradio and Hugging Face Spaces

License

[Your chosen license]