How habeeboss/voicerecognition works
This is a Streamlit audio speaker-classification app. A user uploads an audio file, selects a model (or random selection), and requests classification; the app extracts audio features, applies the corresponding scaler and model, then displays a speaker label. The four model/scaler pairs are distinct runtime inference options. Relationships shown are those visible in the sampled app; no README documentation was provided.
Main entry points: appStreamlit.py Prediction dispatch · ANN MODEL.keras ANN model · ANN SCALER.joblib ANN scaler · CNN MODEL.keras CNN model
Components 14
User interface
- Upload and results
appStreamlit.py - Audio loading
appStreamlit.py
Inference
- Feature extraction
appStreamlit.py - Prediction dispatch
appStreamlit.py - Speaker labels
appStreamlit.py
Model artifacts
- ANN model
ANN MODEL.keras - ANN scaler
ANN SCALER.joblib - CNN model
CNN MODEL.keras - CNN scaler
CNN SCALER.joblib - RNN model
RNN MODEL.keras - RNN scaler
RNN SCALER.joblib - KNN model
KNN MODEL.joblib - KNN scaler
KNN SCALER.joblib
Other components
- User
Connections 15
- User → Upload and results: uploads and selects
appStreamlit.py - Upload and results → Audio loading: loads audio
appStreamlit.py - Upload and results → Prediction dispatch: requests prediction
appStreamlit.py - Prediction dispatch → Feature extraction: extracts features
appStreamlit.py - Prediction dispatch → ANN scaler: transforms with
appStreamlit.py - Prediction dispatch → ANN model: predicts with
appStreamlit.py - Prediction dispatch → CNN scaler: transforms with
appStreamlit.py - Prediction dispatch → CNN model: predicts with
appStreamlit.py - Prediction dispatch → RNN scaler: transforms with
appStreamlit.py - Prediction dispatch → RNN model: predicts with
appStreamlit.py - Prediction dispatch → KNN scaler: transforms with
appStreamlit.py - Prediction dispatch → KNN model: predicts with
appStreamlit.py - Prediction dispatch → Speaker labels: maps predictions
appStreamlit.py - Upload and results → Speaker labels: displays label
appStreamlit.py - Upload and results → User: shows result
appStreamlit.py