How sheelabhadra/emergency-vehicle-detection works
This repository is a collection of research implementations for detecting emergency and alarm sounds from audio, rather than one unified application. Its user-facing workflow is to supply audio, extract signal or learned features, classify emergency presence, and inspect predictions or plots. It preserves three distinct approaches: pitch periodicity (Paper-1), MFCC/ANN classification (Paper-2), and noisy-environment alarm detection (Paper-3), alongside Paper-3’s broader audio-analysis toolkit. Paper-2 internals and several Paper-3 runtime links are documented but not excerpted, so those connections rely on README statements or are omitted.
Main entry points: audioFeatureExtraction.py Feature extraction · audioAnalysis.py Analysis commands · audioSegmentation.py Segmentation · audioBasicIO.py Audio I/O
Components 20
Data acquisition
- Data generation
Data-aquisition/dataGen.py
Pitch detection
- Pitch runner
Paper-1/find_pitch.py - CAMDF pitch
Paper-1/pitch_detection.py - Pitch plot
Paper-1/pitch_detection.py
MFCC classifier
- MFCC training
Paper-2/em_detection.py - Clip prediction
Paper-2/test_audio_file.py - ANN model
Paper-2/model2.h5
Alarm analysis
- Alarm training
Paper-3/em_detection.py - Audio prediction
Paper-3/test_audio_file.py - Live analysis
Paper-3/real_time_test.py - Trained model
Paper-3/model3.h5 - Audio I/O
Paper-3/pyAudioAnalysis/audioBasicIO.py - Feature extraction
Paper-3/pyAudioAnalysis/audioFeatureExtraction.py - Model utilities
Paper-3/pyAudioAnalysis/audioTrainTest.py - Segmentation
Paper-3/pyAudioAnalysis/audioSegmentation.py - Analysis commands
Paper-3/pyAudioAnalysis/audioAnalysis.py - Visualization
Paper-3/pyAudioAnalysis/audioVisualization.py - Feature helpers
Paper-3/pyAudioAnalysis/utilities.py
Other components
- Audio analyst
- Audio recordings
Connections 26
- Audio analyst → Audio recordings: provides
README.md - Audio recordings → Pitch runner: analyzes
Paper-1/find_pitch.py - Pitch runner → CAMDF pitch: runs
Paper-1/find_pitch.py - CAMDF pitch → Pitch plot: plots
Paper-1/pitch_detection.py - Audio recordings → MFCC training: trains on
README.md - MFCC training → ANN model: trains
README.md - Audio recordings → Clip prediction: predicts
README.md - Clip prediction → ANN model: loads
README.md - Audio recordings → Alarm training: trains on
README.md - Alarm training → Audio I/O: imports
Paper-3/em_detection.py - Alarm training → Feature extraction: imports
Paper-3/em_detection.py - Audio recordings → Audio prediction: predicts
README.md - Audio prediction → Trained model: uses
README.md - Audio recordings → Live analysis: analyzes live
- Analysis commands → Audio I/O: imports
Paper-3/pyAudioAnalysis/audioAnalysis.py - Analysis commands → Feature extraction: imports
Paper-3/pyAudioAnalysis/audioAnalysis.py - Analysis commands → Model utilities: imports
Paper-3/pyAudioAnalysis/audioAnalysis.py - Analysis commands → Segmentation: imports
Paper-3/pyAudioAnalysis/audioAnalysis.py - Analysis commands → Visualization: imports
Paper-3/pyAudioAnalysis/audioAnalysis.py - Feature extraction → Audio I/O: imports
Paper-3/pyAudioAnalysis/audioFeatureExtraction.py - Feature extraction → Feature helpers: imports
Paper-3/pyAudioAnalysis/audioFeatureExtraction.py - Segmentation → Feature extraction: imports
Paper-3/pyAudioAnalysis/audioSegmentation.py - Segmentation → Model utilities: imports
Paper-3/pyAudioAnalysis/audioSegmentation.py - Segmentation → Audio I/O: imports
Paper-3/pyAudioAnalysis/audioSegmentation.py - Visualization → Feature extraction: imports
Paper-3/pyAudioAnalysis/audioVisualization.py - Visualization → Model utilities: imports
Paper-3/pyAudioAnalysis/audioVisualization.py