<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"><head><meta http-equiv=Content-Type content="text/html; charset=utf-8"><meta name=Generator content="Microsoft Word 15 (filtered medium)"><style><!--
/* Font Definitions */
@font-face
{font-family:"Cambria Math";
panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0in;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
span.EmailStyle18
{mso-style-type:personal-reply;
font-family:"Calibri",sans-serif;
color:windowtext;}
.MsoChpDefault
{mso-style-type:export-only;
font-family:"Calibri",sans-serif;}
@page WordSection1
{size:8.5in 11.0in;
margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]--></head><body lang=EN-US link=blue vlink=purple style='word-wrap:break-word'><div class=WordSection1><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p><div style='border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0in 0in 0in'><p class=MsoNormal>> <b>On Behalf Of </b>Dave Sill via extropy-chat<br><b>Subject:</b> Re: [ExI] Google AI low-resolution photo enhancement<o:p></o:p></p></div><p class=MsoNormal><o:p> </o:p></p><div><div><div><p class=MsoNormal>On Tue, Sep 7, 2021 at 5:30 PM BillK via extropy-chat <<a href="mailto:extropy-chat@lists.extropy.org">extropy-chat@lists.extropy.org</a>> wrote:<o:p></o:p></p></div></div><div><blockquote style='border:none;border-left:solid #CCCCCC 1.0pt;padding:0in 0in 0in 6.0pt;margin-left:4.8pt;margin-right:0in'><p class=MsoNormal><br>This is really remarkable! It should be a big help to improve low<br>quality photos of criminals caught in the act.<o:p></o:p></p></blockquote><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>>…There's a big difference between taking a low-res image and generating a realistic-looking high-res image from it, and taking a low-res image and recovering lost information from it. This is the former. It'll be great for cleaning up old/bad photos. It'll be useless for zooming in on photos of perps or their license plates. -Dave<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>Dave I agree with the notion of enhancing photos of perps but not so fast on the license plates. Those are two different things.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>You may recall a Hollywood movie with Special Agent Jack Ryan using enhanced remote imagery to figure out if it is the bad guy they are looking for or someone else. Ryan has to make the decision to assassinate him using a drone based on his best guess from enhanced imagery, knowing he mighta just authorized killing an innocent person.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>The physics hipsters hated that part, knowing that what they did with image enhancement is impossible.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>However… under some circumstances, when you already know a limited set of what the hypothetical perfect image should look like, there is some digital enhancement possible and useful. So I agree with your notion you can’t identify a perp, because you don’t know what she should look like under ideal conditions.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>But let us look at the license plate notion, shall we? A license plate has a finite and manageable set of theoretical possibilities. Given sufficient computational resources, a fuzzy image of a license plate can be enhanced in ways the eye cannot do. It can reduce the set of possible plate numbers and states down to a small enough number (such as a few dozen) that each one can be checked and accounted for. Often there is only one of the few dozen candidates which match the car they are registered to hang on.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>So… I would say you are right on the image enhancing tech being useless for perps, only partially right on the plates. Takes a lotta computing resources, but that is plentiful in our times when we have all this computational power originally purchased for bitcoin mining but now outdated for that purpose. It would be practically free to distribute the computing task to hundreds or thousands of amateur crimefighters, give each a subset of the skerjillions of possibilities to look thru, see which ones come back with the best possible matches, bag the sleazy perp from her car’s plates.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>Then of course let her go, because nearly everything is functionally legal in California, but hey, we caught her anyway.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>spike<o:p></o:p></p></div></div></div></div></body></html>