1.4. Using this book#

The recommend starting point is the website https://speechprocessingbook.aalto.fi/. This gives very easy access to most content in the book.

To try out the code-examples and to get full access to all interactive demonstrations, we offer the following options:

  1. Directly from the website with code-examples and interactive components through Mybinder or Colab. (Disclaimer: Mybinder is often slow and on Colab some notebooks do not currently work.) To use this feature, click on the rocket-shaped logo on top of the page.

  2. Download the book from GitHub Speech-Interaction-Technology-Aalto-U/itsp. Run JupyterLab on a local server.

Benefits of running the code yourself include:

  • You can change parameters and code to see how that changes behaviour.

  • You can use your own sound files (such as recording of your own voice) in both code-examples and interactive visualizations.

1.4.1. Using a local server#

While it is not necessary for most readers, to get comprehensive access to all content, you can run a local Jupyter-server to access this book. To make this work, follow the steps below.

  1. Download the book from GitHub Speech-Interaction-Technology-Aalto-U/itsp.

  2. Install Anaconda (recommended) or some other packing-service to get access to Python.

  3. It is recommended (but might not be required) to configure Anaconda by adding the following channels.

    conda config --add channels conda-forge
conda config --add channels pytorch

    Then these channels should appear in your .condarc -file together with default.

  4. Install the following packages

    conda install matplotlib scipy numpy jupyter jupyterlab torch torchaudio ipython ipywidgets

  5. In the folder where the book is stored, run JupyterLab by

    jupyter lab &

  6. Open the file index.md and starting reading.

1.4.2. Exporting to pdf or epub#

To read the material on a mobile device or to print out the whole thing on paper, it can be beneficial to export the whole book to a pdf or epub file. We do not officially support this export (=this might not work), but this is the approach that has been reported to work for some people.

  1. Download the book from GitHub Speech-Interaction-Technology-Aalto-U/itsp.

  2. Install Anaconda (recommended) or some other packing-service to get access to Python.

  3. It is recommended (but might not be required) to configure Anaconda by adding the following channels.

    conda config --add channels conda-forge
conda config --add channels pytorch


    Then these channels should appear in your .condarc -file together with default.

  4. Install the following packages

    conda install matplotlib scipy numpy jupyter jupyterlab torch torchaudio ipython ipywidgets jupyterbook


    Observe that in difference to section Local server, here we need to install also the package jupyterbook.

  5. Go to the folder where you have downloaded the git-repository and compile the whole book into a single html-file by

    jb build . --builder singlehtml

  6. Install Pandoc and use it to convert the html to pdf or epub.

These instructions were collected from the discussion at github that also contains some alternative approaches.