Numpy

NumPy is the fundamental package for scientific computing in Python. It is a library for the Python programming language, adding support for large, multi-dimensional arrays and matrices, along with a large collection of high-level mathematical functions to operate on these arrays.

SciPy

The SciPy library is one of the core packages that make up the SciPy stack. It provides many user-friendly and efficient numerical routines such as routines for numerical integration and optimization.

pandas

Software library written for data manipulation and analysis in Python. Offers data structures and operations for manipulating numerical tables and time series.

Matplotlib

Matplotlib is a Python 2D/3D plotting library which produces publication quality figures in a variety of hardcopy formats and interactive environments across platforms.

scikit-learn

The higher level probability algorithms for machine learning. If you know the rules for dealing with your data then you will want something lower level. If you want the computer to learn the rules for you and give you probabilistic answers then this library is useful. This requires study of metaparameters to understand if you are getting a more correct answer than not. Don't reinvent the wheel !

pydub

High level API for the manipulation of an audio file.

ffmpeg

FFmpeg is the leading multimedia framework, able to decode, encode, transcode, mux, demux, stream, filter and play pretty much anything that humans and machines have created. e.g.) this is used to encode and decode mp3 files.

Numpy and Matplot

• conda create --name speech python=2.7 (Creating a virtual environment called speech with python (2.7))
• source activate speech (Activating the virtual environment: speech)
• conda install numpy (Installing numpy package)
• conda install matplotlib (Installing matplotlib package)
• conda install pydub=0.9.0 (Installing pydub)
• sudo apt-get install ffmpeg (Installing ffmpeg)
• python mp_plot.py (Running an example - wav_plot.py)
• source deactivate (Deactivating the virtual environment: speech)

Yaafe: audio feature extraction

• conda install --channel https://conda.anaconda.org/Yaafe yaafe (yaffe installation)

librosa

• conda install -c conda-forge librosa=0.5.1

Anaconda

• CUDA needs to be installed
• conda create --name speech python=2.7 (Creating a virtual environment called speech with python (2.7))
• source activate speech (Activating the virtual environment: speech)
• conda clean --all
• conda install -c anaconda tensorflow-gpu=1.1.0 (cudatookit, cudnn, numpy, ..., of course, tensorflow-gpu)

Tips

• CUDA info: /usr/local/cuda/samples/bin/x86_64/linux/release/deviceQuery
• CUDNN version: /usr/local/cuda/targets/x86_64-linux/include/cudnn.h