audiofeatures 0.1.6

Extract MFCC, spectral, and pitch features from audio files

audiofeatures

audiofeatures is a lightweight Python library for extracting features from .wav audio files using librosa. It can extract MFCCs, spectral features, and pitch for machine learning or audio analysis tasks.

---

Features

• Extract MFCC mean and standard deviation (13 coefficients)
• Compute Zero Crossing Rate (ZCR)
• Compute RMS energy
• Compute Spectral features: centroid, bandwidth, rolloff, contrast
• Compute pitch mean and standard deviation
• Returns results as a pandas DataFrame

---

Installation

pip install audiofeatures

---

Usage

Extract audio features from a single file

from audiofeatures import extract_audio_features

df = extract_audio_features("sample.wav")
print(df.head())

---

Example Output

The extract_audio_features() function returns a pandas DataFrame with columns like:

mfcc_mean_1, mfcc_mean_2, ..., mfcc_std_13,
zcr, rms, spec_centroid, spec_bandwidth, spec_rolloff, spec_contrast,
pitch_mean, pitch_std

---

Dependencies

• librosa
• numpy
• pandas

---

License

MIT License

---

Author

Dhyan Sudheer