Hello, I'm here to seek out a community of other #Disabled people and accomplices.

I'm working to build a non-profit tool to make alt text easier, and I want to connect with folks who can help guide that project toward a useful start.

I've built a proof of concept, but getting it further will take building community. I'm #Autistic, and networking does not come easy to me. Boosting this is an immense help.

More information at https://alt-text.org

Alt-Text.Org

A library of alt text written by humans, in progress. Stay tuned for a roadmap.

@hannah
awesome initiative! You may wish to use some hashtags though as The fedeverse is really big on those and if you don't add some your reach will be severely limited. my favorite way of finding new ones is using the search field for common terms and seeing how many people have posted on them recently.
someone good ones include:
"AltText" "ImageDescriptions" "blind" "VisuallyImpaired" "accessibility" "A11Y" "access" "inclusion" and "disability"
Good luck! And thank you.
@hannah I am disabled in a way that makes a text often difficult as my hands are disabled and text to speech doesn't always work well. I would love a way to make adding alt text easier.
@geographile Can I ask what inputs you use? I want to make sure the user interface we build can work well with controls other than the status quo.
@geographile @hannah It sounds as if you to really enjoy the art of composing ALT, but I want you to know that — if, for whatever reason, you cannot generate ALT yourself — you can always tag your post #ALT4Me, which is monitored by a group of volunteers who'll reply with a description or transcription as needed.
@oedipusnj @geographile That's very appreciated! I generally don't struggle to write it, and I have a database of alt text for my art when I repost it. Knowing that that's out there makes me very happy!
@hannah @geographile ALT4Me has been a real benefit here & at Twitter. We've got it running at Buesky, too, where there is a feed for a ALT4me requests & a separate one for ALT4you submissions.

@hannah @geographile PS: The ALT4Me info was primarily aimed at @geographile & anyone else in a similar situation

I know who you are from your great work at Twitter, Hannah! Many a day I was rate limited because I had invoked the alt text util bot too many times!

@hannah I probably can't help in any way, but I loved your use of the word 'accomplices' rather than 'allies'. Splendid!
@hannah @lacouvee An interesting and much needed project. I like the idea of being able to add alt text to someone else’s image when I boost it rather than adding a reply with alt text (which I do when I accidentally boost an image without it).
@annewalk @lacouvee There's a lot of work to be done hammering out the details, but I think it could be really useful.
@hannah Interesting project. I'm part of a team that writes alt text for astronomical images from Hubble and JWST. Sharing that alt text is more difficult than sharing the images, so I can see the need for a tool that makes it easier.
@kellylepo Your alt text is incredible! The idea grew from designing and building a database of my digital art to keep them next to their alt text, among other things.
@hannah Love this idea! Solidarity, from one disabled autistic dev to another 

@hannah

Two instances that might be of interest to you are:

https://disabled.social/
https://kind.social/

disabled.social

A place for people who are chronically ill, mentally ill, disabled, and friends/families/allies to come together, meet, share knowledge and random banter, and just about anything else.

Mastodon hosted on disabled.social
@hannah
Sounds like a great idea! A lot of times it overwhelms me because I want it to describe all the details and be interesting but that takes a lot more time and effort than "cat on a chair".

@hannah
Hi, tell us more about this.

Unfortunately the site is on Google servers, and we block Google.

@dsfgs Yes, I'm afraid GCP is what I'm familiar with from work, and so it's where I hosted things.

The idea is a library of alt text, keyed by image, offering a public API for usage in a variety of ways. I designed and built the proof of concept alone, and like any proof of concept it set aside some questions, so first priority of the next iteration is getting more input from potential users about usage and other considerations.

Happy to dive into more detail if you'd like.

@hannah Hi ! I like the initiative, but I have a lot of questions, and I didn't find a page that would explain them.

  • How will the images be matched ? By file hash ? Or with an algorithm that would find similarities ? Will it work with animated images as well ?
  • Do you plan to save several alt texts for the same image (Different languages/idioms, but also different contexts with the same image usage ?)

@MarcBrillault The details are intentionally a little vague on that first page, because one of the first steps is revisiting design decisions made when building the proof of concept.

As for the system as it stands:
- Texts are keyed by a SHA256 hash of the image pixels, w/ additional fuzzy matching using an algorithm here: https://github.com/alt-text-org/image-algo-testing
- As it stands it will only be able to examine the first frame of animated images
- It supports multiple texts, same lang, tagged with their lang.

GitHub - alt-text-org/image-algo-testing: Testing various image matching algorithms' performance on the Pinecone vector DB

Testing various image matching algorithms' performance on the Pinecone vector DB - alt-text-org/image-algo-testing

GitHub

@hannah Thanks for your answers.
If I understand well, there will be no possibilities to have several different texts per language ?

If that is so, I feel this will be a problem, alt the same image can be used in several contexts, even memes. Do I want to use an image of the office because of the character, what is said, or the face's expression ? Do I want to speak about the actor ? their outfit ?
It would be irrelevant to list all of those, so having a choice between several possible alts would be better.

And that opens lots of questions, like "how will the alt texts be presented ?" "How can we manage to limit the number of possible alt texts for one image ?"

I've thought a lot about this myself, and never found the willpower to do something for now.

@MarcBrillault The system as it stands stores text keyed by (image_hash, user_id, lang), so while a single user couldn't store multiple texts for the same (image_hash, lang), there could be multiple texts available for that image. It's not perfect, but it was good enough for a proof of concept.

I have possible answers for your questions, and so far I've spent a lot of time thinking about all of them, but one of my core goals is re-exploring the design decisions I made without discussion.

@hannah ok, no worries then. Thanks for your patience !
@hannah I like your project! I'm #Disabled, too (amputee and vision). I'd like to help how I can, though my tech stack of choice is wildly different (#Perl/#jQuery). Let me know how I can help!
@hannah I don't need ALT text yet, but one day I might. Great timeworthy project. I'll follow and contribute where I can, thanks for the info and I wish you and the project success 🙏
@hannah boosted to support you in your endeavors.
hey @weirdwriter this might be up your alley!
@hannah I’m not at all tech savvy but it is so great to see community building and togetherness 💚
@hannah
Hi there! I’m deaf but have cochlear implants, so I somewhat hear. I’ve also worked in IT as a coder (PHP, HTML, JS etc) for my whole adult life, so around 20 years. Hope I can find a way to take part in this.
@prettyhuman I'd like that too! I'll make announcements here as we identify places folks can contribute.

@hannah it’d be wonderful if alt-text could be translated with ease from other languages into one’s own language

For me, I’d find that useful for foreign-language cartoons which might have the speech bubble text in each frame in a language I don’t understand, I’d be able to understand and enjoy the cartoon

@u0421793 The plan is definitely to tag alt text submitted with a language. I don't think translation will be in the cards, at least in the start of the project, unfortunately.
@hannah could AI help with your project? Like something that explains what an image contains or what is happening in the image. If I'm not wrong it will be possible with ChatGPT 4
@stefh AI image descriptions have been around a while, and are getting better, but they still fall down in the general case. They can be a great place to start editing for a human, and I hope to offer them as such.
@hannah don't use google please. they are evil.
@ecosurrealism If I can figure out a way to do this off GCP with a tiny volunteer team, assuming I can find ops-y folks to help, I will. I'm not sure I can.
@hannah hosting, docker development, count me in :) we like to take this beautiful projects away from google systems.

@ajeremias The needs are less simple than just needing hosting. Some of the technologies the proof of concept relied upon in GCP don't have a 1:1 option elsewhere, among other issues.

Given the set of functionality that's needed and likely too resource-intensive to build for a small team, we're tied to some cloud provider. I'm not excited to work with any of them.

@hannah can you explain me better? Why you say so? I offer that help specifically. If you dont need is also ok..

@ajeremias Sure, off the top of my head this has two major reliances on GCP:

1. Backend DB - We need a horizontally scalable backend database that supports multiple indices on a single table. Maintaining that is hard, and it goes 0-100 as far as cost to host yourself.
2. Rate Limiting - We'll need some fairly tight rate limiting to constrain abuse. Building that is *hard*

Aside from those there are also other places where I just don't think we'll have the support to build it ourselves.

@hannah Would you be interested in hearing my two immediate quick thoughts on the structuring of the project?
@gaditb I'm happy to, with the proviso that I'm intentionally being a bit vague in describing its structure. One of the immediate steps is reassessing the design decisions I made in the first go-round.

@hannah
(Possibly you already thought of this.)

- this has user-submitted content being spread to an amplified audience. It's going to need active moderation (that scales with its use) and can't be just a passive programmatically-automated tool.

- alt-text is contextual, to its use and to its audience. It will probably need to have multiple options present for selection.

... but maybe not if social media use in particular is the goal, like if you're doing a webpage/news story you're putting in enough effort that alt-text-ing it yourself isn't significantly more?
... but maybe still, people use things differently in social media, there isn't always one single use of an image.
... but maybe there's a notably Most Common use.
(... but should the fact that an image can get used in multiple languages be considered?)

@gaditb Those are all definite considerations going into the next phase of the project, and they are definitely high on the list of things that will shape our design. I have long-standing ideas for how to handle those issues, but one of my priorities in seeking community is having folks to provide feedback once we're ready to put forward a more concrete plan around dealing with them.
@hannah Sounds good. I figured it was likely you had thought of these already, but they were significant enough and wide-impacting enough that I figured it was worth possibly redundently checking in the low probability they had been missed.

@hannah Like, you can ban slurs/censored versions of slurs ("k*ke", e.g.) algorithmically, sure -- which might limit the category of images you can provide the service for, like some images do legitimate contain slurs and are shared and described in detail for e.g. documenting hate-speech, but it's reasonable to just say "that's out of scope because it's too sensitive, people will need to alt-text this themselves" --

but more to the point, you can neither ban (e.g., these are examples and not trying to be a full or equal list) pictures of black people or pictures of Jews, nor ban the words "gorilla" or "vermain".

There will need to be at some level manual moderation against that.

@gaditb That is a very specific concern I share, and one that I'm still thinking through the best approach for and hoping to get feedback once we have a more concrete plan. It's complicated by the fact that, for a variety of reasons, we ideally wouldn't store the images themselves.
@hannah I just finished an annoying conversation about the differences between compleance and access regarding alt text with some folks, so am uniquely aware of this conversation and am super interested however the provided url is currently not working