Software resources

Software for feature extraction

• Librosa is a Python package that includes acoustic feature calculation
• MIRtoolbox is a MATLAB toolbox for music information retrieval, but has functions for feature extraction that can also be used in speech analysis
• OpenSMILE is a C++ based feature extraction library

Speech recognition tools

• Kaldi is a C++ state of the art library for automatic speech recognition that includes both GMM and DNN based
• Hidden Markov Model Toolkit is a C-based tool for automatic speech recognition with GMMs
• CMU Sphinx
• Exemplar-based processing tools for robust speech recognition (MATLAB)

Speaker recognition

• ALIZÉ is a C++ opensource platform for speaker recognition

Paralinguistic tools

• OpenEAR is C++ library for paralinguistic analysis of speech

Speech separation and enhancement tools

• NMF-based MATLAB tool for separation of known sources
• FASST, a toolbox for source separation  (both MATLAB and C++)

Databases