Blind and Visually Impaired Community
↑
2
↓
Chrome Extension to Provide Alt Attributes to Image Tags (self.Blind)
submitted by dmuzzin
Hello,
Some of you might've saw my post about a month ago regarding accessibility features you wish your iPhone had, but if you didn't here's a brief background on what I'm working on. I'm currently in the senior design course for my university's computer science degree and this course is centered around designing accessible software. More specifically, the class's focus is to design software for a high school girl named India, who is visually impaired. India frequently attends our class to discuss our projects and answer any questions about her experience using technology, but I figured it would be best to reach out to a larger community for feedback on our project. My group eventually decided our project was going to be a Google Chrome extension that identifies images on a web page that are missing alt attributes, sends the images off to various Computer Vision API's to describe or transcribe them, judges whether the description is good or not, then adds an alt attribute with a sufficient description or transcription from the API's we are using.
We have our own ideas about how to implement this / what features we want to have, but we have many questions regarding internet browsing with a screen reader who's answers will help us refine our project to be as helpful as possible.
Which screen reader do you use, and when it encounters an image with no alt attribute or with an empty alt attribute what does it do?
Would you much rather have a prompt that asks you to transcribe or describe all images on a web page without alt attributes, or would you prefer a case-by-case basis where when you reach an image on the page, you are prompted with whether or not you want the alt attribute to be added? Note that we would have to modify how the screen reader works so that it recognizes that there is an image there, which I assume can be done but might need to be implemented on a screen reader by screen reader basis.
Since the National Federation of the Blind v. Target case, it seems as if all retailers at the very least have an alt attribute for the main image of the product on the page. I was just curious about what the rest of the online shopping experience is like for the visually impaired. Are the text titles of the products usually enough? Would replacing an alt attribute like "Samsung 60 Inch LED TV" with a description like "Family watching movie on Samsung TV" so that a company's marketing efforts can reach consumers who use screen readers be of any value? In summary, what more do you wish you had regarding images on online shopping websites?
Memes! For something that's as simple as text on top of an image it's absurd to me how there hasn't been good software designed to make memes accessible to the visually impaired. Do any of you already have some sort of software that transcribes memes for you? Given that part of the humor is based on image underneath the overlaying text is meme transcription a service you would be interested in? Would a simple but informative web page that describes the actual image of many popular reoccurring memes be helpful so that you can understand the context of the meme text?
I'm sure I'll come up with more questions in the future, but if you have any other thoughts or suggestions be sure to throw them in the comments.
Thank you!
Some of you might've saw my post about a month ago regarding accessibility features you wish your iPhone had, but if you didn't here's a brief background on what I'm working on. I'm currently in the senior design course for my university's computer science degree and this course is centered around designing accessible software. More specifically, the class's focus is to design software for a high school girl named India, who is visually impaired. India frequently attends our class to discuss our projects and answer any questions about her experience using technology, but I figured it would be best to reach out to a larger community for feedback on our project. My group eventually decided our project was going to be a Google Chrome extension that identifies images on a web page that are missing alt attributes, sends the images off to various Computer Vision API's to describe or transcribe them, judges whether the description is good or not, then adds an alt attribute with a sufficient description or transcription from the API's we are using.
We have our own ideas about how to implement this / what features we want to have, but we have many questions regarding internet browsing with a screen reader who's answers will help us refine our project to be as helpful as possible.
Which screen reader do you use, and when it encounters an image with no alt attribute or with an empty alt attribute what does it do?
Would you much rather have a prompt that asks you to transcribe or describe all images on a web page without alt attributes, or would you prefer a case-by-case basis where when you reach an image on the page, you are prompted with whether or not you want the alt attribute to be added? Note that we would have to modify how the screen reader works so that it recognizes that there is an image there, which I assume can be done but might need to be implemented on a screen reader by screen reader basis.
Since the National Federation of the Blind v. Target case, it seems as if all retailers at the very least have an alt attribute for the main image of the product on the page. I was just curious about what the rest of the online shopping experience is like for the visually impaired. Are the text titles of the products usually enough? Would replacing an alt attribute like "Samsung 60 Inch LED TV" with a description like "Family watching movie on Samsung TV" so that a company's marketing efforts can reach consumers who use screen readers be of any value? In summary, what more do you wish you had regarding images on online shopping websites?
Memes! For something that's as simple as text on top of an image it's absurd to me how there hasn't been good software designed to make memes accessible to the visually impaired. Do any of you already have some sort of software that transcribes memes for you? Given that part of the humor is based on image underneath the overlaying text is meme transcription a service you would be interested in? Would a simple but informative web page that describes the actual image of many popular reoccurring memes be helpful so that you can understand the context of the meme text?
I'm sure I'll come up with more questions in the future, but if you have any other thoughts or suggestions be sure to throw them in the comments.
Thank you!
fastfinge 1 points 6y ago
> sends the images off to various Computer Vision API's to describe or transcribe them
Yes, please. I'm not sure how I feel about information about every page I visit getting sent somewhere, though. Plus this sounds expensive. Why not put a context menu item like "transcribe images on this page" when a page is right clicked, and then just find images without alt text on that one page and try and describe them. I don't need this 99% of the time. But the 1% of the time I need it, I really do need it.
> Memes! For something that's as simple as text on top of an image it's absurd to me how there hasn't been good software designed to make memes accessible to the visually impaired.
Usually, once I have the caption from a particular meme, I can find information about it on websites like know your meme. Getting the caption can be quite a painful process, though. So even if you could just get a reliable transcription of the text, that'd be enough for me to search Google and get the context. Bonus if you could somehow identify popular memes, and link to the relevant page. But that's just saving me a Google search, so not a big deal.
Yes, please. I'm not sure how I feel about information about every page I visit getting sent somewhere, though. Plus this sounds expensive. Why not put a context menu item like "transcribe images on this page" when a page is right clicked, and then just find images without alt text on that one page and try and describe them. I don't need this 99% of the time. But the 1% of the time I need it, I really do need it.
> Memes! For something that's as simple as text on top of an image it's absurd to me how there hasn't been good software designed to make memes accessible to the visually impaired.
Usually, once I have the caption from a particular meme, I can find information about it on websites like know your meme. Getting the caption can be quite a painful process, though. So even if you could just get a reliable transcription of the text, that'd be enough for me to search Google and get the context. Bonus if you could somehow identify popular memes, and link to the relevant page. But that's just saving me a Google search, so not a big deal.
dmuzzin [OP] 1 points 6y ago
Thanks for the reply! In regards to your first point, the way our extension currently works is that when your browser begins to load the web page, it first asks you whether or not you would like the images to be transcribed rather than just doing it automatically. Also if you have any feedback on the online shopping experience for the visually impaired and would be willing to message it to me that would be awesome.
fastfinge 1 points 6y ago
The thing about the online shopping experience is that I just ignore the images entirely. I do all my online shopping at costco and Amazon, and both those sites offer good product descriptions. But at the same time, I have no idea what the images on those sites show, so it's hard for me to say that having them described wouldn't add anything. But I don't feel like I'm missing information I need to help me decide if I'm going to buy or not. Plus, every online shopping site I've ever been to has disclaimers like "product may not be exactly as shown". So that makes me wonder even more about the worth of the images.
This nonprofit website is run by volunteers.
Please contribute if you can. Thank you!
Our mission is to provide everyone with access to large-
scale community websites for the good of humanity.
Without ads, without tracking, without greed.
scale community websites for the good of humanity.
Without ads, without tracking, without greed.
©2023 HumbleCat Inc • HumbleCat is a 501(c)3 nonprofit based in Michigan, USA.