How foamyfsky/noisy_pattern_detect works
Noise Sentinel is a local acoustic-burst detector for one laptop. Its main workflow accepts generated, uploaded, or microphone audio; extracts acoustic features; detects and merges burst events; optionally stores them; and presents event summaries and time-based forecasts and anomaly scores in a Streamlit dashboard. SQLite event history feeds the time models, while CSV export supports taking records out of the app. The README documents the workflow; some dashboard and export wiring is only partially excerpted, so uncertain links are omitted.
Main entry points: app.py Streamlit dashboard · storage.py Event persistence and export · pipeline.py Analysis pipeline · signal_processing.py Acoustic features
Components 14
Dashboard
- Streamlit dashboard
app.py - Event summaries
noisy_pattern_detect/reporting.py
Audio analysis
- Demo audio
noisy_pattern_detect/demo.py - WAV and mic capture
noisy_pattern_detect/audio_capture.py - Acoustic features
noisy_pattern_detect/signal_processing.py - Burst detection
noisy_pattern_detect/detection.py - Detection and model settings
noisy_pattern_detect/config.py - Analysis pipeline
noisy_pattern_detect/pipeline.py
Event records
- Event persistence and export
noisy_pattern_detect/storage.py - SQLite event database
- CSV event export
Time models
- Time-of-day intensity
noisy_pattern_detect/models.py - Poisson forecast and anomaly
noisy_pattern_detect/models.py
Other components
- Laptop user
Connections 17
- Laptop user → Streamlit dashboard: uses
README.md - Streamlit dashboard → Demo audio: generates
app.py - Streamlit dashboard → WAV and mic capture: loads or records
app.py - Streamlit dashboard → Acoustic features: extracts features
app.py - Streamlit dashboard → Burst detection: detects events
app.py - Analysis pipeline → Acoustic features: extracts features
noisy_pattern_detect/pipeline.py - Analysis pipeline → Burst detection: detects events
noisy_pattern_detect/pipeline.py - Analysis pipeline → Event persistence and export: optionally stores
noisy_pattern_detect/pipeline.py - Streamlit dashboard → Event persistence and export: loads and stores
app.py - Event persistence and export → SQLite event database: reads and writes
noisy_pattern_detect/storage.py - Event persistence and export → CSV event export: exports
noisy_pattern_detect/storage.py - Streamlit dashboard → Event summaries: summarizes events
app.py - Streamlit dashboard → Time-of-day intensity: requests intensity
app.py - Streamlit dashboard → Poisson forecast and anomaly: requests scores
app.py - Time-of-day intensity → Poisson forecast and anomaly: supplies rate
noisy_pattern_detect/models.py - Detection and model settings → Acoustic features: configures
noisy_pattern_detect/signal_processing.py - Detection and model settings → Burst detection: configures
noisy_pattern_detect/detection.py