xbob.db.verification.filelist 1.3.5

Modify your setup.py and download from PyPI

That is the easiest. Edit your setup.py in your satellite package and add the following entry in the install_requires section (note: ... means whatever extra stuff you may have in-between, don’t put that on your script):

install_requires=[
  ...
  "xbob.db.verification.filelist",
],

Proceed normally with your bootstrap/buildout steps and you should be all set. That means you can now import the namespace xbob.db.verification.filelist into your scripts.

Modify your buildout.cfg and download from git

You will need to add a dependence to mr.developer to be able to install from our git repositories. Your buildout.cfg file should contain the following lines:

[buildout]
...
extensions = mr.developer
auto-checkout = *
eggs = bob
       ...
       xbob.db.verification.filelist

[sources]
xbob.db.verification.filelist = git https://github.com/bioidiap/xbob.db.verification.filelist.git
...

Creating file lists

The initial step for using this package is to provide file lists specifying the world (train), development and evaluation set to be used by the biometric verification algorithm. The following files need to be created:

• For training:

1. world file, with default name train_world.lst, and default subdirectory norm. It is a 2-column file with format:

   filename client_id

2. two (optional) world files, with default names train_optional_world_1.lst and train_optional_world_2.lst, and default subdirectory norm. The format is the same as for the world file. These files are not needed for the most of the face recognition algorithms, hence, they need to be specified only if the algorithm uses them.

• For enrollment:

1. two model files for the development and evaluation set, with default name for_models.lst and default subdirectories dev and eval respectively. They are 3-column files with format:

   filename model_id client_id

• For scoring:

1.a. two probe files for the development and evaluation set, with default name for_probes.lst and default subdirectories dev and eval respectively. These files need to be provided only if the scoring is to be done exhaustively, meaning by creating a dense probe/model scoring matrix. They are 2-column files with format:

filename client_id

1.b. two score files for the development and evaluation set, with default name for_scores.lst and default subdirectories dev and eval respectively. These files need to be provided only if the scoring is to be done selectively, meaning by creating a sparse probe/model scoring matrix. They are 4-column files with format:

filename model_id claimed_client_id client_id

2. two (optional) files for t-score normalization for the development and evaluation set, with default name for_tnorm.lst and default subdirectories dev and eval respectively. They are 3-column files with format:

   filename model_id client_id

3. two (optional) files for z-score normalization for the development and evaluation set, with default name for_znorm.lst and default subdirectories dev and eval respectively. They are 2-column files with format:

   filename client_id

The summarized structure of the base directory (here denoted as basedir) containing all the files should be like this:

basedir -- norm -- train_world.lst
       |       |-- train_optional_world_1.lst
       |       |-- train_optional_world_2.lst
       |
       |-- dev -- for_models.lst
       |      |-- for_probes.lst
       |      |-- for_scores.lst
       |      |-- for_tnorm.lst
       |      |-- for_znorm.lst
       |
       |-- eval -- for_models.lst
               |-- for_probes.lst
               |-- for_scores.lst
               |-- for_tnorm.lst
               |-- for_znorm.lst

Protocols and file lists

When you instantiate a database, you have to specify the base directory that contains the file lists. If you have only a single protocol, you could specify the full path to the file lists described above as follows:

>>> db = xbob.db.verification.filelist.Database('basedir/protocol')

Next, you should query the data, WITHOUT specifying any protocol:

>>> db.objects()

Alternatively, if you have more protocols, you could do the following:

>>> db = xbob.db.verification.filelist.Database('basedir')
>>> db.objects(protocol='protocol')

When a protocol is specified, it is appended to the base directory that contains the file lists. This allows to use several protocols that are stored in the same base directory, without the need to instantiate a new database. For instance, given two protocols ‘P1’ and ‘P2’ (with filelists contained in ‘basedir/P1’ and ‘basedir/P2’, respectively), the following would work:

>> db = xbob.db.verification.filelist.Database('basedir')
>> db.objects(protocol='P1') # Get the objects for the protocol P1
>> db.objects(protocol='P2') # Get the objects for the protocol P2

Note that if you use several protocols as explained above, the scoring part should be defined in the same way for all the protocols, either by using for_probes.lst or for_scores.lst. This means that at the time of the database instantiation, it will be determined (or specified using the use_dense_probe_file_list optional argument), whether the protocols should use the content of for_probes.lst or for_scores.lst. In particular, it is not possible to use a mixture of those for different protocols, once the database object has been created.