CAMOMILE : Collaborative Annotation of multi-MOdal, multI-Lingual and multi-mEdia documents

CAMOMILE (2012-2016) is a CHIST-ERA project selected for the call “From data to new knowledge” (D2K).

The project involves 6 participants from 4 countries:

• LIMSI (Orsay, France),
• IMMI (Orsay, France),
• LIG (Grenoble, France),
• UPC (Barcelona, Spain),
• LIST (Esch-sur-Alzette, Luxembourg) and
• ITU (Istanbul, Turkey).

It is funded by ANR, FNR, Tubitak and Mineco.

In the context of the CAMOMILE project, we developed a collaborative annotation framework for multi-modal, multi-lingual and multi-media data:

• A documentation with all the routes available on this server can be found at http://camomile-project.github.io/camomile-server.
• The source code of the camomile server is distributed under MIT open source license at https://github.com/camomile-project/camomile-server and is also provided in a Docker image allowing an easy installation.
• Along with the server, Python and JavaScript clients embedding the REST API into native language objects have also been developed and are distributed in sibling GitHub repositories.
• More details about the framework here.

CAMOMILE organized the The 2015 Multimodal Person Discovery in Broadcast TV task during MediaEval 2015 with support from ELRA and INA. It was followed with the 2016 Person Discovery edition. More details are available in the MediaEval 2015 proceedings and MediaEval 2016 proceedings.

The CAMOMILE framework was presented at Interspeech 2015 Show&Tell or LREC 2016. See the list of 50 publications related to the project.