Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
flipboard.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
Welcome to Flipboard on Mastodon. A place for our community of curators and enthusiasts to inform and inspire each other. If you'd like to join please request an invitation via the sign-up page.

Administered by:

Server stats:

1.2K
active users

flipboard.social: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#imagedescription

0 posts0 participants0 posts today
Public
Jupiter Rowland
@Joseph Meyer
When you read exceptional alt text, do you ever compliment its author? What is the epitome of alt text, either in general terms or using a specific example?

I'd really like to know that myself, also to up my own game further and always stay way ahead of image description quality requirements.

I mean, I've learned a lot about describing images in and for the Fediverse over the last two years. But I guess I can still learn something new, even if I think I already take care of everything, even if the technical possibilities I have here on Hubzilla for describing images surpass those on Mastodon by magnitudes.

Maybe, if I learn something new from those who reply, I can weave it into the image descriptions for a series of images that I've been working on since late last year (the descriptions, not the images which are ready to go).

Alt text sometimes merely explains what I am viewing; other times it draws my attention to special details in a photo that I would have otherwise missed.

I never explain in alt-text. I do always explain a whole lot because I always have to explain a whole lot. For my original images, it takes me over 1,000 characters alone to explain where an image was made.

But I only ever give explanations in the long, detailed image descriptions that go into the post text body (in addition to shorter and purely visual descriptions in the alt-texts).

Or if there's no additional long image description in the post itself which is the case for my meme posts, I still supply enough explanation in the post text body (still not in the alt-text) for just about everyone in the Fediverse to understand them without having to look anything up themselves. If I can link to external information, e.g. KnowYourMeme for the template I've used, I do so. If I can't, I write the missing explanations right into the post myself.

#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta
hub.netzgemeinde.euImage descriptions in the FediverseI have learned a lot about describing images according to Mastodon's standards, and I want to share my knowledge, but I haven't learned enough
#cwimagedescriptionmeta
Replied in thread
Public
Jupiter Rowland
@David Mitchell :CApride:
Mostly, just imagine you’re telling your friend over the phone about image you’re looking at and what they would need to know.


Let's just say I'm a bit critical about that because, in my opinion, it doesn't work in the Fediverse.

Jupiter Rowland schrieb den folgenden Beitrag Fri, 04 Oct 2024 23:30:02 +0200

You can't describe images in Fediverse posts like over the phone

Allegedly, a "good" advice for image descriptions is always to describe images like you'd describe them to someone on a landline phone.

Sorry, but that's non-sense. At least for anything that goes significantly beyond a real-life cat photo.

If you describe an image through a phone, you describe it to one person. Usually a person whom you know, so you've at least got a rough idea on what they need described. Even more importantly, you can ask that person what they want to know about the image if you don't know. And you get a reply.

If you describe an image for a public Fediverse post, you describe it to millions of Fediverse users and billions of Web users. You can't know what they all want, nor can you generalise what they all want. And you can't even ask one of them what they need described before or while describing, much less all of them. In fact, you can't ask at all. And yet, you have to cater to everyone's needs the same and throw no-one under a bus.

If I see a realistic chance that someone might be interested in some detail in one of my images, I will describe it. It won't be in the shorter description in the alt-text; instead, it will be in the long description which I've always put directly into the post so far, but whose placement I'm currently reconsidering. If something is unfamiliar enough to enough people that it requires an explanation, I will explain it in the long description.

Right now, only meme posts are an exception. They don't need as much of a visual description as long as I stick to the template, and a poll has revealed that people do prefer externally linked third-party explanations over my own ones blowing the character count of the post out of proportion. This is the one time that I can safely assume that I actually know what most people want.

@accessibility group @a11y group

#Long #LongPost #CWLong #CWLongPost #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta #Inclusion #A11y #Accessibility

CC: @Monstreline @Claire (sometimes Carla) @qurly(not curly)joe

#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta #QuotePost #QuoteTweet #QuoteToot #QuoteBoost
#quoteboost
Replied in thread
Public
Jupiter Rowland
@Anna Maier I don't know what constitutes a "good" example in your opinion, but I've got two examples of how bad AI is at describing images with extremely obscure niche content, much less explaining them.

In both cases, I had the Large Language and Vision Assistant describe one of my images, always a rendering from within a 3-D virtual world. And then I compared it with a description of the same image of my own.

That said, I didn't compare the AI description with my short description in the alt-text. I went all the way and compared it with my long description in the post, tens of thousands of characters long, which includes extensive explanations of things that the average viewer is unlikely to be familiar with. This is what I consider the benchmark.

Also, I fed the image at the resolution at which I posted it, 800x533 pixels, to the AI. But I myself didn't describe the image by looking at the image. I described it by looking around in-world. If an AI can't zoom in indefinitely and look around obstacles, and it can't, it's actually a disadvantage on the side of the AI and not an unfair advantage on my side.

So without further ado, exhibit A:

This post contains
  • an image with an alt-text that I've written myself (1,064 characters, including only 382 characters of description and 681 characters of explanation where the long description can be found),
  • the image description that I had LLaVA generate for me (558 characters)
  • my own long and detailed description (25,271 characters)
The immediate follow-up comment dissects and reviews LLaVA's description and reveals where LLaVA was too vague, where LLaVA was outright wrong and what LLaVA didn't mention although it should have.

If you've got some more time, exhibit B:

Technically, all this is in one thread. But for your convenience, I'll link to the individual messages.

Here is the start post with
  • an image with precisely 1,500 characters of alt-text, including 1,402 characters of visual description and 997 characters mentioning the long description in the post, all written by myself
  • my own long and detailed image description (60,553 characters)

Here is the comment with the AI description (1,120 characters; I've asked for a detailed description).

Here is the immediate follow-up comment with my review of the AI description.

#Long #LongPost #CWLong #CWLongPost #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta #AI #LLaVA #AIVsHuman #HumanVsAI
llava.hliu.ccLLaVA
#humanvsai
Replied in thread
Public
Jupiter Rowland
@sunflowerinrain @Tarnport From what I've read, a digital photograph is considered the default. So for brevity reasons, it must not be mentioned.

Any other media must be mentioned, whether it's a painting, a screenshot from a social media app, a scanned analogue photograph, a flowchart, a CAD blueprint, a 3-D rendering or whatever.

But an alt-text must never start with "Image of", "Picture of" or "Photo of". That's considered bad style and a waste of characters and screen-reading time. If the medium is not mentioned, digital photograph falls into its place as a default.

#Long #LongPost #CWLong #CWLongPost #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta
hub.netzgemeinde.euNetzgemeinde/Hubzilla
#cwimagedescriptionmeta
Replied in thread
Public
Herzleid

@nikolar #Alt4You #ImageDescription
A majestic photo of a magpie atop her nest at the top of a telephone pole. The nest is made from intricately woven twigs and branches, then decorated by a lot of shiny glittering items, chains, pendants, necklaces, key chains, bracelets, something that looks like a pocket watch… It's the epitome of magpie opulence.

Replied in thread
Public
Jupiter Rowland
@Alt Text Hall of Fame @David Bloom Yes.

Explanations, or any other information available neither in the image nor in the post text, must never ever go into the alt-text. That's because not everyone can access alt-text. And to those who can't access alt-text, any information exclusively available in alt-text is inaccessible and therefore lost.

#AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta
hub.netzgemeinde.euNetzgemeinde/Hubzilla
#cwimagedescriptionmeta
Replied in thread
Public
Jupiter Rowland
@Christina
Please make it descriptive: tell us about colours, what the person looks like, the background, and anything else you can think of.

I've got a few examples of image posts of my own with very detailed image descriptions. These posts are not on Mastodon, but they are very much in the Fediverse in places that are federated with Mastodon, and they all have actually ended up on Mastodon timelines. So I guess Mastodon's accessibility requirements apply to these places just as well.

Would you say they're too descriptive? Or would you even say they aren't descriptive enough yet, and that there are important details missing?

One example is (content warning: eye contact) this post with two images, my most recent one. It is not available on mastodon.online. These images also show a digital avatar, so in a sense, I've described a person.
  • The first image has exactly 1,500 characters of alt-text, 990 are short image description, 509 explain where the long description can be found.
  • The second image has 1,499 characters of alt-text, 989 are short image description, 509 explain where the long description can be found.
  • The long descriptions have almost 11,000 characters in the common preamble, over 2,800 characters specifically for the first image and almost 6,600 characters specifically for the second image because it has more surroundings. All in all, the long descriptions have almost 21,000 characters. It took me eight hours to research for and write them. The visual description of the avatar plus necessary explanations alone takes up almost 7,000 characters.

Another example is this post with one image. You can also find it by searching for the hashtag #UniversalCampus and scrolling to the bottom of the results; there you can see what it looks like on Mastodon.
  • The alt-text is exactly 1,500 characters long, 1,402 of which are the short image description. The note where to find the long description had to be shortened to the point of being half-useless.
  • The long description in the post is over 60,000 characters long. It took me two full days, morning to evening, to research for and write it.
  • I've actually had to limit myself in comparison to earlier image descriptions: There are no longer any detailed descriptions of images within the image, especially not at a higher level of detail than what the images within the image themselves show in-world.
  • This description is outdated because I've used absolute measures rather than measures relative to what people are familiar with, and the descriptions of the colours may not be detailed enough.

#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta #Inclusion #A11y #Accessibility
#accessibility
Replied in thread
Public
Kevin Karhan :verified:

@dandylover1 thanks!

I've shared this post of you because more people need to know that Alt Text and Image Descriptions are not only good but are appreciated by those that rely on them!

  • The strenght of the Fediverse is based upon it's diversity, and that must include accessibility!

After all it's easy for sighted people to dismiss it when in reality it doesn't take away anything from them if they just write a sentence or two per image.

  • Thabks for letting us know!
#AltText#ImageDescription#Accessibility
Public *
semblancestudio

Exploring the impact of #StormÉowyn in the park.

The birds seemed disoriented, still finding their feet and wings the day after. Many trees fallen. 💕 to all who are feeling the impact.

#Nature #NaturePhotography #Glasgow #Scotland

#ImageDescription A Weimeraner dog gazes across a branch-covered path in Pollok park in Glasgow. Two orange construction cones cover the edge of the path, and a fallen tree blocks part of the path ahead. A black Spaniel dog looks on from nearer the viewer.

A Weimeraner dog gazes across a branch-covered path in Pollok park in Glasgow. Two orange construction cones cover the edge of the path, and a fallen tree blocks part of the path ahead. A black Spaniel dog looks on from nearer the viewer.
ExploreLive feeds
About