The contest is over. Thanks to the people participating in the contest. The presentation held at the ICPR in Istanbul can be seen here.

In an increasing number of mobile devices a digital camera is integrated. More and more the recorded video material is put on internet archives like YouTube. From a forensic perspective, recordings of illegal activities and illegal content have special interest. That is, witnesses of illegal activities can themselves be actionable or at least be important witnesses. Production of illegal content is generally a higher offense than possession only. For instance in case of child pornography, for the Dutch law possession is not allowed, but the production results in a higher penalty. Further, any footage can serve as material to test testimonies and confessions. 
 
Essential in these situations is the establishing of the source camera of the video footage, which answers the question of ‘who made the video’ to a great extent. For original footage found on networks and personal hard disks, the problem of source camera determination is doable with state-of-the art methods. The most challenging problem in this respect is the source camera determination for transcoded videos. This happens to videos uploaded to YouTube among others. Upon upload, an additional video codec transcodes the video stream. Moreover, it may also be reformatted to a standard resolution. 
 
In forensics, the set of possible source cameras for a questioned video is mostly unknown. Therefore, a closed-set recognition setting is inappropriate. For the CAMCOM contest, the source camera determination problem is approached as a verification problem. That is, the question to be answered is: ‘is the given camera the source of the questioned video?’

The CAMCOM benchmark will be organized as activity of the International Association for Pattern Recognition (IAPR) technical committee on Computational Forensics TC06. For the biennial International Conference on Pattern Recognition (ICPR) we organized a challenge for source camera recognition. In TC06 we consider peer-review, performance evaluation and benchmarking of algorithms and computational procedures as a fundamental contribution to establish the scientific basis for a forensic discipline and to provide tools supporting the forensic examiner in his / her daily casework.

Each contestant/group can submit an overview paper and working notes paper describing the used methodology and findings. If this is submitted before the regular paper submissions and are accepted after peer review, these papers will be published in the main conference proceedings.