In this little guide, we'll provide you with some technical details we expect for the detector. You should know the overall process of the submission from prior submissions at this point.
This is the same as for the hider or restorer, so for further information have a look there
The in- and output will be again be in the JSON-format. We provide the input to your tool via stdin and expect the output at stdout. Your detector will get a feed as its input. The feed might have been tampered with by a hider, or be an actual clean feed. In both cases expect the input to look like this:
{"feed": ["<article 1>",
"<article 2>", …, "<article 30>"]}
The task for the detector is to assess whether the provided feed contains a
hidden message. In case a feed is believed to do so, the result argument of
the output should be set to true, and false
otherwise. The detector's output is expected to have the following
structure:
{"result": true | false}
Our submission platform will first check against the specification before assessing your results. If the format is correct, we provide you the precision and recall for your detector.
Like in the self-assessment, the hider, and the restorer, there must be a
main.py which can be run with python3 main.py. That
is, without any arguments.