In the framework of the 5th European Cyber Week and the 26th C&ESAR conference, the second CAID (Conference on Artificial Intelligence for Defense) conference will be held. Dedicated to Artificial Intelligence for Defence, it will take place in Rennes on 18 and 19 November 2020. A call for publications has been launched and papers can be submitted before Monday 31 August 2020. The second CAID (Conference on Artificial Intelligence for Defense) conference dedicated to artificial intelligence applied to defense will take place in Rennes on 18 and 19 November 2020, as part of the 5th European Cyber Week (organized by the Cyber Pole of Excellence and its partners from 17 to 19 November) and the 26th C&ESAR conference (organized by the French Ministry of the Armed Forces). Artificial intelligence is a vast field that ranges from neural networks to more traditional statistical approaches (SVM, decision tree forests...) through probabilistic methods (Bayesian networks...), not forgetting datascience. Current advances in artificial intelligence, particularly in deep learning, have led to significant breakthroughs in many civilian applications. These civilian applications can be transposed to the world of defence. In particular, we can cite the following non-exhaustive examples:
  • Unstructured data processing algorithms, such as Detection, Recognition and Identification (DRI) algorithms, which are widely used in the civilian world on RGB images and which can be transposed to other types of images (SAR, Infrared, hyperspectral...), or audio processing algorithms which can have variations for data such as radar, sonar....
  • Anomaly detection technologies that allow, among other things, the exploitation of data when labels are not necessarily available.
In addition, the use of these technologies in the defence sector brings additional challenges. Indeed, the needs in terms of operational safety bring to light problems related to the qualification of these systems:
  • Luring, in particular neural networks or any other attack targeting AI, which can pose security problems in some applications using AI.
  • Technologies allowing to qualify AI, i.e. to provide evidence and guarantees on the functioning of AI whether it is empirical measurements or formal evidence; methodologies/metrics allowing to qualify a learning or test database.
  • Confidentiality of training data, i.e. the possibility or not to trace back to potentially classified training data characteristics.
  • Finally, AI algorithms must also be able to be embedded on compact systems with low power consumption. Thus all technologies allowing to reach this goal (quantification, approximation) and what they imply (performance, robustness, qualification) are of interest.

CAID 2020 Conference

This year, the CAID conference proposes a special session dedicated to the use of Artificial Intelligence and/or Datascience for crisis management. In this framework, applied contributions to the fight against COVID 19 are welcome. The CAID conference is intended to be educational and presents issues related to artificial intelligence to an audience that may be new to this field. It is within this framework that we invite actors from all fields (academic research, industry, defence...) to submit publications related to one of the topics in the field of artificial intelligence and defence.

How to submit

Submission of articles before Monday 31 August 2020 via the following url in PDF format: https://easychair.org/conferences/?conf=caid2020 It should include the title of the paper, the first and last names of the authors and their affiliation, the e-mail address of the main author, an abstract (10 lines max.) and a list of keywords. The paper should be a maximum of 8 pages, A4 format, following the Springer Lecture Notes in Computer Science model. Particular attention will be paid to the pedagogical quality of the submitted article. It is preferable to submit a paper that presents the context and issues well, rather than a paper that is too pointed and not very accessible. Authors will be notified of the acceptance or rejection of their paper on Friday, October 15. The final version of the article with reviewers' comments must be provided by Friday, October 30.

Programme Committee

  • Alain DRONIOU (DGA)
  • Guillaume QUIN (MBDA)
  • Alexis OLIVEREAU (CEA)
  • Christophe MEYER (THALES)
  • Adrien CHAN HON TONG (ONERA)
  • Judicaël MENANT (DGA)
Translated from Appel à contributions pour la CAID (Conference on Artificial Intelligence for Defense) qui se tiendra à Rennes

Our articles will then appear first in Google Top Stories.