Jeffrey D. Stark

was on a bit of accessibility training today and the sighted person was demoing a screen reader and talking about people who are blind. He started talking about how most screen reader users don't see and then how all blind folk are naturally slower at doing stuff online. If I hadn't objected, I don't think they would have clarified that what they meant was that on sites and so on that aren't accessible, or designed inclusively they would be slower. However; this is the kind of thing that gets under my skin and I may over react to but is a legitimate issue in terms of framing people with disabilities when giving #a11y and training on accessibility testing

So I just got the pre-release Agiga Echo smart glasses - designed for and by the blind Monday. Have been playing with them since.. Thought I might share my 1st impressions:

Pros:
- The glasses are super light and comfortable especially compared to glasses like the meta raybands / oakleys. They fit my head which I was worried about because hats and other items often don’t
- The buttons, controls and case are well designed
- Scene and image descriptions seem solid and live AI mode also works well
- Text reading works well
- No “daddy knows best” filtering of what the AI will read [e.g. license plates, medical notes and other things some Ais like Meta will refuse to rea through the glasses

Cons:
- Beep alerts: Could you guys not have found a slightly less abrasive beep for the case? It is on par with a fire detector’s beep. It beeps every time the glasses are taken out or put in the case, for battery levels and other things. My family reacted really negatively to the sound. They cringe every time I use it and ask me if I can turn it off or do something about it. Can this be turned off, disabled or changed? If not, this means others around will react just as poorly too.
- The TTS and speech items are clear and pleasanteven when spe up
- Audio sound on the glasses is good but not on part with some of the other mainstream glasses on the market
- The Android setup process has bugs
- The live AI mode has complained about speed and other connectivity issues
- I’m not completely convinced Wifi only is the best option but will stay impassive until I’ve had more time with the unit
- Cloud AI is slow and more on device functionality should be explored

So far, seems a solid solution.

For perspective, I’ve played with most of the glasses from this and previous generations including several metas, envision, orcam, etc.

My met experience was written up here: https://blog.blackspheretech.com/?p=485

#a11y #blind #AI

@matt@toot.cafe Hey Matt, long time no speak. I am reaching out because I am working on an interdepartmental working group to develop a desktop software accessibility testing methodology for the government of canada. We are starting with Windows. We ran into some complex discussions about the intersection of OS APIs/AT. They asked if anyone knew someone with direct experience with screen readers on windows who might be willing to share their insights during 1 of our meetings. (These are primarily compsci professionals who work in digital accessibility in the GC) Given your background in writing adaptive technology I thought immediately of you and was wondering if you would consider it. We are focusing on the technical intersection between accessibility APIs, adaptive technology behavior, and desktop software design. Specifically, they were looking to ask questions around:
How software exposes information to the operating system through accessibility APIs such as UI Automation and IAccessible2
The reality of how adaptive technology like JAWS and NVDA intercept and translate API calls
Semantic mapping between native Windows controls and custom UI elements
How desktop applications populate the accessibility tree and handle focus changes
Differences in how native Windows applications handle off-screen content compared to web browsers
The relationship between system themes, performance, and accessibility API communication
Methods to verify the complete chain from software through API to adaptive technology to the user
We would benefit greatly from your insights on how AT interprets what the API provides and where the most common breakdown points occur.

Thank you for considering this.