sigmoid.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A social space for people researching, working with, or just interested in AI!

Server stats:

594
active users

#ISCC

0 posts0 participants0 posts today
Europe Says<p><a href="https://www.europesays.com/1956155/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="">europesays.com/1956155/</span><span class="invisible"></span></a> EU Scrutinizes Fraud in Certification of Biofuels <a href="https://pubeurope.com/tags/europe" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>europe</span></a> <a href="https://pubeurope.com/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a></p>
DW Innovation<p>With digital seals, broadcasters can make the provenance of their content transparent and visible to users. How do those markers work, and what are their limitations?</p><p>New blog post by our lab colleague Daniela Späth:</p><p><a href="https://innovation.dw.com/articles/digital-seals-trust-news-media" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">innovation.dw.com/articles/dig</span><span class="invisible">ital-seals-trust-news-media</span></a></p><p><a href="https://mastodon.social/tags/C2PA" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>C2PA</span></a> <a href="https://mastodon.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> <a href="https://mastodon.social/tags/blockchain" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>blockchain</span></a> <a href="https://mastodon.social/tags/provenance" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>provenance</span></a> <a href="https://mastodon.social/tags/trust" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>trust</span></a></p>
Gerrit Gragert<p><span class="h-card" translate="no"><a href="https://openbiblio.social/@awinkler" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>awinkler</span></a></span> <span class="h-card" translate="no"><a href="https://mastodon.social/@titusz" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>titusz</span></a></span> <span class="h-card" translate="no"><a href="https://mastodon.social/@Posth" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Posth</span></a></span> If we want to compare whole collections of images, I think we need first of all a kind of ground truth data to learn about the threshold in the <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> of images to say these are different digitisations of the same object or if it's better to do some pre-processing on the image properties (contrast, brightness, colors etc.) to get better results. But we will never get 100% similarity while comparing different scans from the same physical object.</p>
Gerrit Gragert<p><span class="h-card" translate="no"><a href="https://openbiblio.social/@awinkler" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>awinkler</span></a></span> <span class="h-card" translate="no"><a href="https://mastodon.social/@titusz" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>titusz</span></a></span> I compared the two images in the <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> Playground (even though <span class="h-card" translate="no"><a href="https://mastodon.social/@Posth" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Posth</span></a></span> notes that it's still experimental) and came up with a match in the content code of 82%. If I cut away the black border on one of the postcards, it's 87%. IMHO these are already good values for comparing different digitized objects, which also correspond to my tests with different works.</p>
Alexander Winkler<p>So now I've tried comparing two identical postcards (<a href="https://commons.wikimedia.org/wiki/File:Berlin-Kreuzberg_Postkarte_021.jpg" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">commons.wikimedia.org/wiki/Fil</span><span class="invisible">e:Berlin-Kreuzberg_Postkarte_021.jpg</span></a>, <a href="https://berlin.museum-digital.de/object/141040" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">berlin.museum-digital.de/objec</span><span class="invisible">t/141040</span></a>) using the <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> hash and got the following comparison map ... not quite sure what to make of it. Maybe <span class="h-card" translate="no"><a href="https://mastodon.social/@titusz" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>titusz</span></a></span> can help again? Ideally, I'd use the hash to check an third party image collection against wiki commons objects enriched with ISCC statements.</p>
Gerrit Gragert<p><span class="h-card" translate="no"><a href="https://mastodon.social/@Posth" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Posth</span></a></span> <span class="h-card" translate="no"><a href="https://openbiblio.social/@Lambo" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Lambo</span></a></span> Daher bin ich gespannt auf die Darstellung der Use-Cases. Wir können uns auch gerne irgendwann mal zusammen setzen und bibliothekarische Use-Cases diskutieren. Interessant für uns wäre z.B. ein <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> über ein gesamtes Werk (=Konglomerat von Scans / Images). Grundsätzlich sehe ich weiterhin ein großes Potential.</p>
Gerrit Gragert<p>vergleichen wir (1) bzw. (2) mit Seite 13 (statt Seite 12) von (3) liegt die Ähnlichkeit im Content bei -6%. Mit dem <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> lässt sich offenbar tatsächlich eine Aussage über die Ähnlichkeit verschiedener Digitalisate von ein und dem selben Werk aus unterschiedlichen Institutionen treffen.<br>4/</p>
Gerrit Gragert<p>Im Sommer hatten <span class="h-card" translate="no"><a href="https://openbiblio.social/@Lambo" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Lambo</span></a></span> und ich schon einmal über den <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> gebloggt <a href="https://blog.tib.eu/2024/07/05/der-international-standard-content-code-iscc-warum-bibliotheken-archive-und-museen-ihn-nutzen-sollten/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">blog.tib.eu/2024/07/05/der-int</span><span class="invisible">ernational-standard-content-code-iscc-warum-bibliotheken-archive-und-museen-ihn-nutzen-sollten/</span></a></p><p>Für eine ganz praktische Anwendung hier einmal hier einmal drei unterschiedliche Digitalisate eines Werks (1) <a href="http://mdz-nbn-resolving.de/urn:nbn:de:bvb:12-bsb00078903-8" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">http://</span><span class="ellipsis">mdz-nbn-resolving.de/urn:nbn:d</span><span class="invisible">e:bvb:12-bsb00078903-8</span></a> (2) <a href="http://mdz-nbn-resolving.de/urn:nbn:de:bvb:12-bsb00078904-3" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">http://</span><span class="ellipsis">mdz-nbn-resolving.de/urn:nbn:d</span><span class="invisible">e:bvb:12-bsb00078904-3</span></a> (3) <a href="http://resolver.staatsbibliothek-berlin.de/SBB00030BBF00000000" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">http://</span><span class="ellipsis">resolver.staatsbibliothek-berl</span><span class="invisible">in.de/SBB00030BBF00000000</span></a> <br>Ich habe im <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> playground jeweils Seite 12 verglichen. Folgend die Ergebnisse:<br>1/</p>
TIB<p>ISO 24138:2024 - wie bitte? <span class="h-card" translate="no"><a href="https://openbiblio.social/@Lambo" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Lambo</span></a></span> und <span class="h-card" translate="no"><a href="https://openbiblio.social/@EzellaGarnie" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>EzellaGarnie</span></a></span> von der <span class="h-card" translate="no"><a href="https://openbiblio.social/@stabi_berlin" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>stabi_berlin</span></a></span> erklären im neuesten Post des <a href="https://openbiblio.social/tags/TIBBlog" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TIBBlog</span></a> instruktiv und anschaulich den International Standard Content Code (<a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a>)und warum Bibliotheken, Archive und Museen diesen nutzen sollten 👉 <a href="https://blog.tib.eu/2024/07/05/der-international-standard-content-code-iscc-warum-bibliotheken-archive-und-museen-ihn-nutzen-sollten/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">blog.tib.eu/2024/07/05/der-int</span><span class="invisible">ernational-standard-content-code-iscc-warum-bibliotheken-archive-und-museen-ihn-nutzen-sollten/</span></a></p>
Gerrit Gragert<p>Have you ever heard of the new International Standard Content Code <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> identifier? For those of you thinking "What's that?", <span class="h-card" translate="no"><a href="https://openbiblio.social/@Lambo" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Lambo</span></a></span> and I have written an article explaining the <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> how it differs from classic <a href="https://openbiblio.social/tags/PID" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PID</span></a> like <a href="https://openbiblio.social/tags/DOI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DOI</span></a> and why it's important for <a href="https://openbiblio.social/tags/GLAM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GLAM</span></a> institutions.</p><p><a href="https://blog.tib.eu/2024/07/05/the-international-standard-content-code-iscc-why-libraries-archives-and-museums-should-use-it/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">blog.tib.eu/2024/07/05/the-int</span><span class="invisible">ernational-standard-content-code-iscc-why-libraries-archives-and-museums-should-use-it/</span></a></p><p>CC <span class="h-card" translate="no"><a href="https://mastodon.social/@Posth" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Posth</span></a></span></p>
Gerrit Gragert<p>Schon mal was vom neuen Identifier International Standard Content Code <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> gehört? Für alle, die jetzt denken "Wasn das?" haben <span class="h-card" translate="no"><a href="https://openbiblio.social/@Lambo" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Lambo</span></a></span> und ich einen Artikel geschrieben, der sowohl den <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> erklärt als auch beschreibt, was ihn von klassischen <a href="https://openbiblio.social/tags/PID" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PID</span></a> wie <a href="https://openbiblio.social/tags/DOI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DOI</span></a> unterscheidet und warum er für <a href="https://openbiblio.social/tags/GLAM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GLAM</span></a> Institutionen wichtig ist.</p><p><a href="https://blog.tib.eu/2024/07/05/der-international-standard-content-code-iscc-warum-bibliotheken-archive-und-museen-ihn-nutzen-sollten/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">blog.tib.eu/2024/07/05/der-int</span><span class="invisible">ernational-standard-content-code-iscc-warum-bibliotheken-archive-und-museen-ihn-nutzen-sollten/</span></a></p>
Gerrit Gragert<p>Got the print version of the information and documentation of the International Standard Content Code <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a>. A new identifier for digital content and objects. That's the library way.</p>
Lambert Heller<p>"Recently, the <a href="https://openbiblio.social/tags/C2PA" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>C2PA</span></a> adopted the <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> as one identifier that is among its list of authoritative soft binding algorithms in the C2PA model." </p><p>Todd goes on describing very well the strength of <a href="https://openbiblio.social/tags/PID" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PID</span></a>|s, and how PIDs and ISCC will complement each other.</p><p>Introducing the Newest ISO Identifier Standard: <span class="h-card" translate="no"><a href="https://social.niso.org/@tac_NISO" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>tac_NISO</span></a></span> (Todd Carpenter, NISO Executive Director) on ISCC, DOI and DIDs.</p><p>via <span class="h-card" translate="no"><a href="https://mastodon.social/@Posth" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Posth</span></a></span> <br>cc <span class="h-card" translate="no"><a href="https://openbiblio.social/@EzellaGarnie" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>EzellaGarnie</span></a></span> </p><p>Try out the ISCC yourself: <a href="https://huggingface.co/spaces/iscc/iscc-playground" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">huggingface.co/spaces/iscc/isc</span><span class="invisible">c-playground</span></a></p>
Nicol Wistreich<p>Huge news (in the standards &amp; decentralised media &amp; <a href="https://social.coop/tags/DID" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DID</span></a> world). The <a href="https://social.coop/tags/ISO" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISO</span></a> has approved <a href="https://social.coop/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> as an official standard: <a href="https://iscc.io/iso-publishes-new-standard-enabling-content-transparency-iso-24138/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">iscc.io/iso-publishes-new-stan</span><span class="invisible">dard-enabling-content-transparency-iso-24138/</span></a></p><p>Historically media was identified in big semi-private databases like ISBN/ISSN.</p><p>Online, YouTube invested $100m+ on Content ID to identify what's uploaded, and recognise material that's already been flagged. </p><p>But <span class="h-card" translate="no"><a href="https://mastodon.social/@Posth" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Posth</span></a></span> &amp; Titusz Pan in <a href="https://iscc.codes" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">iscc.codes</span><span class="invisible"></span></a> (<a href="https://social.coop/tags/ISO24138" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISO24138</span></a>!) have turned the media file itself into an identifier, far beyond trad MD5/hashing.</p>
Gerrit Gragert<p>What great news! The International Standard Content Code <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a> finally became <a href="https://openbiblio.social/tags/ISO" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISO</span></a> 24138. I am quite sure this is a groundbreaking change for <a href="https://openbiblio.social/tags/Scholcom" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Scholcom</span></a> <a href="https://openbiblio.social/tags/GLAM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GLAM</span></a> and digital cultural heritage in general. <a href="https://openbiblio.social/tags/PID" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PID</span></a> <a href="https://iscc.io/iso-publishes-new-standard-enabling-content-transparency-iso-24138/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">iscc.io/iso-publishes-new-stan</span><span class="invisible">dard-enabling-content-transparency-iso-24138/</span></a> via <span class="h-card" translate="no"><a href="https://openbiblio.social/@Lambo" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Lambo</span></a></span></p>
Lambert Heller<p>I'm proudly involved in this effort to create an open, reproducible standard for content identification, which right now has been acknowledged by ISO: <a href="https://iscc.io/iso-publishes-new-standard-enabling-content-transparency-iso-24138/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">iscc.io/iso-publishes-new-stan</span><span class="invisible">dard-enabling-content-transparency-iso-24138/</span></a> This is not only important for the creative industríes, but for scholarly communcation and cultural heritage data as well! In addition to registered names (PIDs), we need modern, smart means to identify content, and chunks of content. No matter in which context, whether hidden or not. <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a></p>
Lambert Heller<p>On the ISCC Playground you can test the experimental Semantic Code of the <a href="https://openbiblio.social/tags/ISCC" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ISCC</span></a>. It is a compact binary code based on deep learning vector embeddings, which allows matching and comparison of semantic characteristics of content files.<br>The Semantic Code is engineered to be robust against a broader range of variations that cannot be matched with the perceptual Image-Code.<br><a href="https://huggingface.co/spaces/iscc/iscc-playground" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">huggingface.co/spaces/iscc/isc</span><span class="invisible">c-playground</span></a></p>