The Web is an amazing repository of information,
but much of this information isn't accessible to blind users.
Although assorted screen readers are readily available,
they may not take advantage of metadata (e.g., HTML tags) on web pages
or web-based image description resources.
Depict provides a platform for the blind
to easily request image descriptions from sighted volunteers.
Depict uses a browser extension and crowd-sourced image descriptions.
The results from this approach could be very accurate and specific,
but there could also be significant latency and scaling issues.
So, let's examine alternative and/or supplemental approaches.
Automated services such as Google's Cloud Vision API
and Microsoft's Project Oxford
could address these limitations,
providing immediate (if limited) descriptions.
So, a blind user could use an automated service to narrow down a set of images,
then call on Depict for a more detailed information.
Most web pages make use of HTML tags to structure the content,
distinguish or emphasize certain text, etc.
However, most screen readers (etc) simply ignore this metadata.
It should be possible to capture this markup
and translates it into a cleaner and simpler form.
For more information, see my Markdown
To be continued...
This wiki page is maintained by Rich Morin
an independent consultant specializing in software design, development, and documentation.
Please feel free to email
comments, inquiries, suggestions, etc!