jameslyons/python_speech_features
Python Speech Features
Python speech features is a signal processing toolkit and library for extracting standard speech recognition features from raw audio signals. It provides computational capabilities to calculate mel-frequency cepstral coefficients, raw and log filterbank energies, and spectral subband centroids for automated speech recognition systems and acoustic analysis.
The library implements audio signal transformations including pre-emphasis filtering, overlapping windowed frame segmentation, Fast Fourier Transform spectral analysis, mel-filterbank projection with configurable filter weights and frequency ranges, and discrete cosine transform cepstral projection.
Features
- Audio Feature Extraction - Computes standard audio representations such as mel-frequency cepstral coefficients and filterbank energies from raw audio.
- Mel-band Energy Extractions - Transforms audio frequency spectrums into perceptually-scaled filterbank energies using configurable mel-triangle filter weights.
- Speech Recognition - Extracts standard audio representations like mel-frequency cepstral coefficients for speech recognition tasks.
- Feature Extraction - Provides a signal processing toolkit for computing cepstral coefficients and filterbank energies for speech recognition.
- Mel-Frequency Cepstral Coefficients - Applies a discrete cosine transform to convert log filterbank energies into cepstral coefficients.
- Fourier Transforms - Converts windowed time-domain audio frames into frequency-domain power spectra using numerical Fourier transforms.
- Spectral Centroids - Computes spectral subband centroids from audio signals to characterize frequency distribution.
- Windowed Segmentation - Segments continuous audio waveforms into overlapping short-time frames using configurable window functions.
- Audio Frequency Analysis - Analyzes audio signals by computing spectral subband centroids to characterize frequency distributions.
- Pre-Emphasis Filters - Applies a high-pass digital filter to raw audio signals to amplify high frequencies.
