Software

Most of the software I develop is open source and is available at GitHub. Recently, I have been coding mostly in Python and Julia. Below are some of the projects to which I contribute:

• SRMRToolbox: a non-intrusive objective metric for speech quality and intelligibility for normal hearing listeners and cochlear implant users. Distributed as a MATLAB toolbox.
• SRMRpy: a Python implementation of the SRMR Toolbox (WIP).
• AuditoryFilterbanks: cochlear and modulation filterbanks from the Auditory Toolbox translated to C++. Includes MEX wrappers.
• DSP.jl: a signal processing library in Julia.
• AuditoryFilters.jl: auditory and modulation filterbanks, based on the Auditory Toolbox by Malcolm Slaney.
• ECOS.jl: a Julia wrapper to the ECOS second-order cone programming solver.
• Boltzmann.jl: Restricted Boltzmann Machines in Julia. I added support for training Deep Belief Networks (DBNs) to this library.
• Keras: a Theano-based Deep Learning library. I helped adding support for HDF5 datasets, saving models to HDF5 files, and general bug fixes.
• mushra-ruby: a Ruby webapp to store results from perceptual listening tests done with mushraJS.
• maracas: a Python library for generating speech datasets with added noise and reverberation.