Vernor Vinge's Rainbows End includes this as a minor early plot point; libraries being fed into giant shredders that digitize and stitch together the scraps on their way to the incinerator.
Wow, fascinating! I hope I’m not the only OCR-minded soul on here that immediately started thinking of pipelines for this one could assemble from all our new DL-enabled toys… it’s a shame that “some American kid” probably doesn’t meet their contracting requirements.
One thing that kinda baffles is that it sounds like the snippets (I bet that German word is fun) haven’t been scanned(/photographed…?) yet. No matter who they contract with, wouldn’t that necessarily be step #1? Or is the scanning the part that needs contracting, and the “e-Puzzler” name was just a lil melodramatic?
For one thing, I feel like it’ll be wildly helpful that they’re on a variety of papers, not challenging like they imply. Once that and bag index have narrowed down neighbor candidates, you’ve got rules and glyphs to go off of, much less recognizable form layouts. It feels like the Vesuvius scrolls but actually tractable…
> Since the project began in 2007, based on contractual obligations, around 6.5 million euros have been paid to Fraunhofer IPK or the Fraunhofer Gesellschaft
> In the fall of 2013, Fraunhofer IPK was able to demonstrate that the software developed by its scientists, the so-called e-Puzzler, works in principle. However, the process proved unsuited for mass production. Overall, the project never achieved the envisaged automated reconstruction rate of at least 80 per cent.
I find the assertion that it was “impossible” to reassemble shredded documents… disappointing - as they seem to be confusing “very difficult with current tools” with “no way of doing it, ever”
I mean, they made sure it was true by burning them all after they decided it could never be done. Phenomenally short-sighted - unless of course there was a reason they wanted those particular documents to never be read.
Don't neglect the possibility that there are people in power who don't want these records to come to light. Either because of direct implication of people or their families in the stasi atrocities, "national embarrassment", etc
Vernor Vinge's Rainbows End includes this as a minor early plot point; libraries being fed into giant shredders that digitize and stitch together the scraps on their way to the incinerator.
Wow, fascinating! I hope I’m not the only OCR-minded soul on here that immediately started thinking of pipelines for this one could assemble from all our new DL-enabled toys… it’s a shame that “some American kid” probably doesn’t meet their contracting requirements.
One thing that kinda baffles is that it sounds like the snippets (I bet that German word is fun) haven’t been scanned(/photographed…?) yet. No matter who they contract with, wouldn’t that necessarily be step #1? Or is the scanning the part that needs contracting, and the “e-Puzzler” name was just a lil melodramatic?
For one thing, I feel like it’ll be wildly helpful that they’re on a variety of papers, not challenging like they imply. Once that and bag index have narrowed down neighbor candidates, you’ve got rules and glyphs to go off of, much less recognizable form layouts. It feels like the Vesuvius scrolls but actually tractable…
> Since the project began in 2007, based on contractual obligations, around 6.5 million euros have been paid to Fraunhofer IPK or the Fraunhofer Gesellschaft
> In the fall of 2013, Fraunhofer IPK was able to demonstrate that the software developed by its scientists, the so-called e-Puzzler, works in principle. However, the process proved unsuited for mass production. Overall, the project never achieved the envisaged automated reconstruction rate of at least 80 per cent.
Well, that was tax-payer's money well spent!
I find the assertion that it was “impossible” to reassemble shredded documents… disappointing - as they seem to be confusing “very difficult with current tools” with “no way of doing it, ever”
I mean, they made sure it was true by burning them all after they decided it could never be done. Phenomenally short-sighted - unless of course there was a reason they wanted those particular documents to never be read.
There's nothing technical to discuss; where's the meat? This 2012 article at least has pictures: https://www.bbc.com/news/magazine-19344978
I speculate the problem is not technical but financial or otherwise; the ML to do this was up to the task at least ten years ago.
This was a DARPA challenge back in 2012 too. Shredded documents seem more of a challenge than torn.
https://cacm.acm.org/news/darpa-shredder-challenge-solved/
Don't neglect the possibility that there are people in power who don't want these records to come to light. Either because of direct implication of people or their families in the stasi atrocities, "national embarrassment", etc
"We simply rehired all Nazi's and Stasi's but here there is a sophisticated technical problem to keep you busy."