Image descriptions on @twitter are great!
But I've been thinking about how an auto-generated description -- that you must edit for acceptance -- could triangulate on good and useful descriptions and an awesome training dataset for machines (and guide for people).
-
-
Have you seen the ImageNet roulette thing going around? I believe most of the images that made up the training set were hand-labeled by humans. Short answer is because humans are terrible and like to label images with cruel stereotypes.
-
Yea, I can see that. Narrowing the training set on twitter would certainly be difficult (e.g Tay bot). But if the set of <(tweet, image), description> examples came from a11y twitter, I'd love to see something like
@HemingwayApp for descriptions because I'm bad at them. - 2 more replies
New conversation -
-
-
two answers: 1. there are small examples of forcing alt text correction, companies generally stop doing it b/c users hate it, they don't want to write alt text at all. (although, FB/Insta both let you correct their auto-gen stuff, if you search for it)
-
2. i'm not 100% sure, but performance of image captioning kind of sucks, so not sure it will result in any better data to do it this way vs. just getting more image captions
- 1 more reply
New conversation -
-
-
Also want to know this. I love writing alt texts, especially going ott describing the context & content. It's a fun exercise. But i've no idea if what i'm doing is of any use - is there an "alt text best practice / wishlist" from people who routinely rely on alt text?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.