I am looking for facial recognition software able to spider the web, particularly the Darknet, for information related to digital images. For example, that can find patterns of keywords related to these images, and identify websites or paper-trails persons may have left behind while going about their activities.

Of particular interest is facial recognition software able to take partial, in-profile, or low-resolution pictures and search for similar results, matching persons of interest to information gathered. Again, I would like to include information that has been obscured or hidden from search engines (to the full extent applicable by all relevant law). It would also be useful to find software that can match distinctive objects from photographs and find these as well.

Does anyone know of software capable of performing some or all of these functions? I currently have access to most openly available Linux distributions, as well as Windows 10, though I have limited access to processing power at present and favour software able to function efficiently and/or utilise cloud-computing resources to complete its allotted task(s).

If necessary I am also willing to write my own PostgreSQL or perhaps even MemSQL database to interact with the facial recognition software and write or customise web-spidering software. So at a push any software capable of outputting usable data based on face-recognition algorithms would still be useful as a recommendation.

  • basically you want to create a graph with relations between darknet users that are naive enough to post (their) face on the profile. why is that ?
    – StefanS
    May 4 '16 at 9:29
  • Not at all. I would not dream of spidering Darknet users in general, as I am sure there are no absolute rules as to the legitimacy of their activities. There is a very specific need and this is a purely information gathering exercise at this stage. May 4 '16 at 9:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.