In the Introduction the authors write "Bystander identification may be especially important when the images record criminal activity, as when hostage takers or child sex abusers photograph their victims." and in the Discussion they note "For now, our findings suggest a novel application of high-resolution photography: for crimes in which victims are photographed, corneal image analysis could be useful for identifying perpetrators," thus framing discussion of their techniques in terms of criminal forensics applied against particularly abhorrent criminals.

However, given the current global reality of pervasive long-memoried and wide-reaching state surveillance as revealed, for example, in Edward Snowden-related articles about the NSA published by Laura Poitras and Glen Greenwald -- the collection of phone record metadata for use in reconstructing social network graphs, in particular -- the authors have missed an opportunity to address the surveillance implications of their research. This is somewhat ironic, given their explicitly calling out the potential for mobile phones to "soon carry >39 megapixel cameras routinely," and the ready access to mobile data reportedly enjoyed by the NSA¹, and presumably other state-sponsored spy agencies.

Such techniques could also be used to produce social graphs of individuals imaged directly and imaged in reflection, and these graphs could be linked to the geographic and time-stamp information that accompany mobile phone cameras. All without the consent of the individuals being so mapped, and regardless of any criminal behavior or lack of criminal behavior.

¹ See for example Jacob Appelbaum's 30C3 presentation on this topic here: http://www.youtube.com/wa...