The App That Launched a Thousand Faces?


Having recently published the system specification from the user point of view, I feel a pressing need to define the software elements that will comprise its whole workings.

Foremost among these will be a database of random pics, to form the backdrop of the familiar picks that the user will identify from each gallery in order to proceed.

This is a recent crop of Conservative members of parliament, and they have literally been cropped by my fair hand to fit a 150 x 150 pixel square.

Aside then from a database of known-to-you and unknown-to-you personages, what we need sooner rather than later is an app into which these images can be dropped like underwear into a washing machine, so as to emerge sized as required.

I am conscious that a US company was recently fined $7 million for harvesting faces from the 'net, but this was using bots whereas my own galleries are lovingly picked and arranged by human hands... which I feel no judge could object to.

Our faces are captured daily and often. In fact the sinister state that has more CCTV cameras per head of population than any other is, in fact... the UK. And I don't recall them ever asking me if they could record my image, inadvertently or otherwise.

First of all then, your honour, I proceed on the basis that I am more than happy to remove images from our database upon request. On the other hand, readers of Vanity Fair (the original London novel and not the mag) will like me be sure most people will actually want to be featured among randomised sets of pics... to the extent that we'd be able to monetise the deal in due course.

Secondly, come July 4th many of the above will disappear forever from the limelight, except when kindled by our own secure access system. And I personally feel that they would like people to be reminded that ~ for a brief moment in time ~ each of them was a contender.

Among the most powerful human emotion is the need not to be forgotten, and my pics will forever serve as tiny headstones... or digitised memorials, your grace?

Prior to many going concerns like Reddit becoming established, or indeed Airbnb, the founders realised much spadework needed doing by themselves in order to kickstart their creations ~ and I feel we shall look back fondly on the days I sat at the kitchen table trimming these portraits to the right size. In doing so too I learned a lot, like even at just 40 x 40 pixels resolution how few data are needed to recognise a face!

Popular posts from this blog

Indian Takeaway

The Digi Yatra facial ID system is being rolled out at Indian airports and though it does not eliminate a boarding pass as reported by the Financial Times, it does eliminate the need to corroborate identities at the gate prior boarding. Users are required to upload a selfie, yet in one survey 30% of them did so without knowing what they were signing up for... which critics suggest is the means by which they could be tracked everywhere and anywhere by the government. My takeaway from this is that to implement a system that involves recognising other people whom you might know ideally involves uploads of the same kind to assemble a bespoke database. Whether it will happen in my lifetime, or indeed at all, is debatable. Unless.
Read more

Digital Commoner

Image
Am not one to obsess over stats, but feedback from the Digital Commons is of special interest to one graduate of geography like myself, as it makes for a snapshot of a kind that prior the internet would barely have been possible. Looking then at a variety of papers and bearing in mind some have had a head start, it is interesting to note that pictures are still worth a thousand words. Of the half-dozen airframes I may have circulated by which means we might be elevated by super-sized drones, the one appearing here has proven by far the most popular... and the only one downloaded by commercial helicopter (and bizarrely, access platform) manufacturers. The runner-up meanwhile is a paper describing how people we might recognise might also be used as a means of secure access to computer systems. Yes, we all recognise Ronald McDonald, but would I recognise your grandmother? The spread around the world reflects our common cultures too. For instance in the US two downloads in San Fran and...
Read more

AI or Ay Ay?

Image
Without the assistance of AI or good old-fashioned Google Lens, can you tell the two apart? Well they are in fact the same Yin Qi who heads up nascent facial ID company MEGVII: https://global.faceid.com/technology/face-recognition-liveness ... except that the portrait on the left is from Tatler magazine and that on the right the South China Morning Post. No prizes for guessing which features on the company website, however, and a relief knowing the guy is at least thirty-six and not fresh out of high-school. Good though that interest in identity theft (and faces) spans the globe.
Read more