this post was submitted on 20 Apr 2024
410 points (96.4% liked)

Memes

45649 readers
1573 users here now

Rules:

  1. Be civil and nice.
  2. Try not to excessively repost, as a rule of thumb, wait at least 2 months to do it if you have to.

founded 5 years ago
MODERATORS
 

image descriptionAn infographic titled “How To Write Alt Text” featuring a photo of a capybara. Parts of alt text are divided by color, including "identify who", "expression", "description", "colour", and "interesting features". The finished description reads “A capybara looking relaxed in a hot spa. Yellow yuzu fruits are floating in the water, and one is balanced on the top of the capybara’s head.”

via https://www.perkins.org/resource/how-write-alt-text-and-image-descriptions-visually-impaired/

top 28 comments
sorted by: hot top controversial new old
[–] davel@lemmy.ml 49 points 6 months ago (2 children)

A capybara in the library with a candlestick.

[–] ElderberryLow@programming.dev 13 points 6 months ago

He was definitely the murderer

[–] superfes@lemmy.world 5 points 6 months ago (1 children)
[–] davel@lemmy.ml 6 points 6 months ago* (last edited 6 months ago) (1 children)
[–] TargaryenTKE@lemmy.world 2 points 6 months ago

A true classic

[–] li10@feddit.uk 27 points 6 months ago (1 children)

Bro I fucking love capybaras so much

10/10 animal, fucking brilliant.

[–] tourist@lemmy.world 11 points 6 months ago

my favourite animal

the carbonara

[–] kibiz0r@midwest.social 23 points 6 months ago (1 children)

I like how “description” is one of the components of the… description.

[–] Viking_Hippie@lemmy.world 4 points 6 months ago

See also: self referential

goes to dictionary entry for "recursion"

[–] airbussy@lemmy.one 15 points 6 months ago (5 children)

Potentially also useful for creating good prompts for AI image generators?

[–] Daxtron2@startrek.website 8 points 6 months ago (1 children)

It's essentially by-hand CLIP, that's how the training data for CLIP came into being, it was descriptive text for images.

[–] Omega_Haxors@lemmy.ml 1 points 6 months ago (1 children)

Explains why it sucks so much shit.

[–] Daxtron2@startrek.website 1 points 6 months ago

CLIP is pretty decent for what it does though

[–] pennomi@lemmy.world 6 points 6 months ago

It’s only useful if the AI was trained on similar prompts. A lot of the anime style ones work best with lists of tags, while the realistic ones work best with descriptions like above.

[–] 9488fcea02a9@sh.itjust.works 5 points 6 months ago* (last edited 6 months ago)

Prompts are just the reverse of image recognition AI tagging stuff.

Alt text is exactly the kind of tedious work that AI would be good at doing, but everyone in the fediverse seems to have a huge hate boner for ANYTHING AI...

Fediverse: write a fucking essay every time you post an image.... But make sure you waste time doing it manually, instead of using AI tools!!!

[–] Blaster_M@lemmy.world 2 points 6 months ago

If you have really detailed image tags, a model trained on them can make great outputs.

[–] Omega_Haxors@lemmy.ml 1 points 6 months ago

We don't do that here.

[–] biptoot@lemmy.today 14 points 6 months ago

This is excellent, very useful for continuing to make images accessible on the fediverse

[–] Gradually_Adjusting@lemmy.world 12 points 6 months ago (1 children)

Me writing alt text: Time is a flat circle. God is a sock.

[–] jcg@halubilo.social 3 points 6 months ago (1 children)

Reminds me of my git commit messages!

[–] Gradually_Adjusting@lemmy.world 2 points 6 months ago

Nothing but crucial intel

[–] toastal@lemmy.ml 6 points 6 months ago

Just as important is “decorative images” where you explicitly leave the alt empty https://www.w3.org/WAI/tutorials/images/decorative/

[–] Omega_Haxors@lemmy.ml 5 points 6 months ago

You know those little [] that appear when you upload an image? You can put alt-text in there.

[–] bjornsno@lemm.ee 1 points 6 months ago (2 children)

Ignorant question: isn't alt text primarily for visually impaired people? If so, what is the point of including info about color?

[–] cypherpunks@lemmy.ml 2 points 6 months ago

Color can provide useful context. For example, in the case of this image, imagine if in a thread about it there was some discussion of the ripeness of the yuzu fruit.

[–] h3rm17@sh.itjust.works 1 points 6 months ago (1 children)

You can also become visually impaired at points other than birth in life, and know colours and stuff

[–] bjornsno@lemm.ee 1 points 6 months ago

That's a very good point!

[–] flashgnash@lemm.ee 1 points 6 months ago

Is this not the kind of thing machine vision/language models would be really good at?