Joe Padfield  0:05  
Okay, thank you very much there. We have a number of things that I haven't seen in the chat. See, okay, so a couple of people suggesting they will give it a go, I'd be very pleased for feedback in relation to what Melissa was saying before about user-testing, Often these things we play with them to build - Oh, so it's already done it, and it works! Excellent! That's what we like to hear! It, it's the idea is trying to make the tools available for people to play with now, and explore them, and then in much makes it much easier for them to express what they eat next. Because at the moment, I think,  based on our discussion is 'how do you start?', 'how do you actually start to engage with these things, with this type of tool?' and the one that was presented before I think Tom mentioned Wax, there are a number of these things that you can start to engage with free services on the web, and start to play with IIIF solutions now, and then becomes a lot easier then to potentially to argue why you want institutional services to do these sorts of things, whether they're internal or external, to actually do things on a larger scale for public engagement or for research. This is the intention here is just to help people start actually working with these things, to have a go. Okay, so are there any more questions... we have one more question: Are there any recommendation for organizations who want to use an existing database of objects to generate IIIF manifests, we have a database of about 1200 objects with photos. Okay, I can't answer that one off the top my head, it is something perhaps we could explore. I would guess it comes down to, if my database is something that doesn't have manifests in it at the moment, it'd be about whether you have an API in there to pull the data out and push it through one of the tools that Tom was mentioning about, to generate manifests from the API data that might come from it. So that would be where I would start, but failing that to jump on the IIIF Slack group, and look at the community, or museums website and start asking people questions and it may well be someone else's also worked with that particular one already. Ah yes, in the chat. It has already been integrated, Tom has an example of it, and that's been done. So this is one of the things you'll find the more you start to explore the IIIF community, the more stuff people have already done. You can build on it, or work with it, and there are a lot of really really good tools out there. Right. Can you explain the role of Cogapp in implementing at The [National] Portrait Gallery? Charlotte Can you speak to this one? Your microphone still off Charlotte.

Charlotte Boland  2:49  
The simple answer will be 'no', from the curatorial side of things, that Cogapp. And there are portrait gallery people on the call so feel free to dive in to the chat. To clarify, but they're the agency that we use to work with. 

Joe Padfield  3:11  
Okay, the next question: Tom mentioned the benefits of IIIF are for reconstructing correspondence Could you elaborate on on other use cases? Tom, do you want to respond to that one?

Tom Crane  3:28  
Yeah, well I mean there are, there are millions of use cases for IIIF, and I'm always thinking of new ones. I mean it kind of, the initial motivation is the, the ability to publish works of cultural heritage (paintings books, maps, etc.), and share them, and not have to reinvent the wheel every time you do that. but the IIIF specification gives us this kind of shared digital space for all that stuff. So, almost there are many, many amazing things that can be done, especially with a national collection that was IIIF-enabled. You know the kind of creative reuse and re-mixing of cultural heritage resources, the ability to bring them into annotation tools and do work on them, to bring them into all sorts of workflows and tools that would be just impossible when they're not standardized. Yeah, I mean, there are there are tons of things. If you hang out on the IIIF slack, that's a good place to, to kind of bounce ideas around, definitely.

Joe Padfield  4:33  
I think one of the original use cases I think IIIF kind of bubbled out the libraries and archives field and it was kind of, I think, medieval manuscripts that's been taken apart into individual pages and then trying to restructure or recreate the digital representation of that complete book was one of the initial presentations but,

Tom Crane  4:51  
In fact, Julian has just pasted into the chat. An example of that, where one institution holds some illuminations that have been ripped out of a manuscript and the other institution holds the manuscript and IIIF can be used to digitally recombine these things. There are some great presentations about IIIF, and just the idea of this notion of using IIIF as digital scissors. The ability to kind of extract bits of things and do things, interesting things with them. Yeah, the Awesome IIIF list that has been linked a few times is a really good place to kind of jump off on for IIIF use cases.

Joe Padfield  5:37  
Another use case I can give you that sort of internal point of view, is that historically we have lots of images in folders within the National Gallery, as I said that were siloed. So essentially, if you wanted to have a conversation someone from a different department of a large, several 100 megabyte image you would copy that and drop it into a different folder that you could then discuss, we end up with multiple duplicates very big, big images all over the place, not necessarily emailed, but that has happened in the past. And you get this, this, this sort of growth in the number of, exponential growth almost. Effectively with the IIIF manifest, you can just share the images that you're working on and that's just a text file. The other thing think of is, if you've got a collection, a IIIF  manifest for your collections, you can then generate new bespoke manifests for exhibitions, for publication plans, for stock presentation ideas. And you don't need to then duplicate or share all these images you just have this little text file, which then reuses them. So it's a massive increase in efficiency for your worried IT departments about storage problems and all sorts of other things. I think it's the use case at the bottom line is, how do I share, join, compare, and reuse images without copying and duplicating them over and over again. And then, as Tom says there's tons of applications for that from there on in. And that's just internally within an institution and then you connect to the outside world and connect institutions together and it just blossoms. It's an extremely flexible and powerful solution. Let's see a few more, oops, let's see if there's another question: Can you describe how exposing your connection to IIIF may enable applications of AI to your images.? Thank you Tom. It depends on whether the AI can ingest, I know for a fact, I don't know if Giles is still on the webinar, Yes, I thought that was a leading question Tom... AI systems can be developed to ingest images through a IIIF manifest to do work and then to push it back out again. So, If you have a set of images that are publicly available and you've got a defined license on them. It makes it a lot easier for you to work with universities or development houses or digital, digital agencies to do AI examination of your images because they can ingest, them work on them, and you don't need to send them hard disks and worry about this that and the other, we've already got it in place. You organize, yes we can share these images for research purposes, and then have discussions with lots of different people. There are certain Oxford Visual Geometry Groups are doing some work on ingesting IIIF manifests to do at the AI based image processing, and subject keywording and a number of other things. And there's a growing number, and that's more and more IIIF manifests, or collections of manifests out there, that have nice we reuse licenses on them so more and more AI approaches will be will be done. Because when you're looking at huge scale. It's hard to do.

Tom Crane  8:47  
So I guess one other advantage there is if you develop say a Machine Learning workflow for, for one set of images based on IIIF, then is obviously applicable, just to other IIIF images because you've, you've done, you've done the hard part. Really you might have to train  different models for different purposes but you've the access to the images is now standardized, so you can reuse all that code you've written for one project and don't use it on the next one.

Joe Padfield  9:14  
It's the beauty of the standard genuinely

Charlotte Boland  9:16  
Joe, there's one answer to the Cogapp question in the chat.

Joe Padfield  9:24  
Excellent, thank you very much. Okay, love to actually receive the chat. These things you click Save chat every now and then. Okay we're reaching the end of our time. It's 31 minutes past 6:00. Do we have any other questions or comments? Would anyone else like to join in? I know it's probably very late or very early for some people if you're still on the call. So thank you very much, we still actually have 136 people still hanging on right to the bitter end. So thank you very much for that.