Bob is a free signal-processing and machine learning toolbox originally developed by the Biometrics group at Idiap Research Institute, Switzerland.

The toolbox is written in a mix of Python and C++ and is designed to be both efficient and reduce development time. It is composed of a reasonably large number of packages that implement tools for image, audio & video processing, machine learning and pattern recognition.

This package is only a place-holder for Bob's Wiki and Bug tracker.

If just want to use Bob's functionalities on your experiments, you are not supposed to install this package on your machine, but rather create your own personalised work environment depending on your needs, by collecting individual sub-packages based on your requirements.

If you are developing Bob packages which are supposed to built along side our nightly build system, please read on.

As per-usual, make sure all external dependencies are installed on your host before trying to compile the whole of Bob. Once all dependencies are satisfied, you should be able to:

$ python bootstrap.py
$ ./bin/buildout

You may tweak the options in buildout.cfg to disable/enable verbosity and debug builds, before you run ./bin/buildout.

You can generate the documentation for all packages in this container, after installation, using Sphinx:

$ ./bin/sphinx-build . sphinx

This shall place in the directory sphinx, the current version for the documentation of the package.

You can run a set of tests using the nose test runner:

$ ./bin/nosetests -sv

You can run our documentation tests using sphinx itself:

$ ./bin/sphinx-build -b doctest . sphinx

Packages must be added using a git submodule:

$ git submodule add https://github.com/bioidiap/bob.foo.bar.git layers/2/bob.foo.bar

Then, update the following files:

1. Add your package to the build:

   $ vim layer2.cfg
   

2. Add your package to the overall documentation:

   $ vim index.rst
   

3. Add your package to be "deploy-tested":

   $ vim requirements2.txt
   

4. Add a row for your package in our Packages list

5. Update the dependency graph:

   $ tools/update_dependency_graphs.sh all
   

   The current dependency graph looks like that:

   

Git submodules work by registering a precise commit hash identifier from each submodule along with their repository locations. If you update the submodule, you have to tell the main module that it now needs to use a new commit identifier. If you don't do so, the main module will continue to checkout the old version of the submodule.

In case a submodule is updated, this package will not automatically update its reference to such database. You need to explicitly do it. To update a submodule, first initialize this package:

$ git submodule init
$ git submodule update

Change to the directory where the submodule you want to update lives:

$ cd layers/2/bob.foo.bar

The submodule repositories added by git submodule update are headless. This means they don't have a branch. So, you must first checkout the branch with the new revision you want to update:

$ git checkout master
$ git pull

Switch back to the root of the package and re-add the submodule:

$ cd ../..
$ git commit -m "Updated bob.foo.bar" layers/2/bob.foo.bar
$ git push