Blind and Visually Impaired Community
↑
7
↓
" So Marx Melencio has taken one of our OAK-D variants and built a visual-assistance device for visually-impaired people. He himself is completely blind. What he is building is so cool. We're so excited to continue to help enable these things with embedded AI and computer vision. " (self.Blind)
submitted by MRXGray
**—** Brandon Gilles, CEO of Luxonis (Colorado, USA)
**SOURCE:** $1
​
**•** " OAK boards were funded within 20 minutes of going live ... the campaign has surpassed its $20K goal by more than $300K in just three days. "
**•** " OpenCV and Luxonis created the OAK ... (almost) plug-and-play eyeballs for gadgets ... on-board AI processing ... don’t have to connect anything to the cloud ... uses USB-C for both power and data ... "
**SOURCE:** $1
​
**•** " OpenCV AI Kit aims to do for computer vision what Raspberry Pi did for hobbyist hardware ... "
**SOURCE:** $1
**SOURCE:** $1
​
**•** " OAK boards were funded within 20 minutes of going live ... the campaign has surpassed its $20K goal by more than $300K in just three days. "
**•** " OpenCV and Luxonis created the OAK ... (almost) plug-and-play eyeballs for gadgets ... on-board AI processing ... don’t have to connect anything to the cloud ... uses USB-C for both power and data ... "
**SOURCE:** $1
​
**•** " OpenCV AI Kit aims to do for computer vision what Raspberry Pi did for hobbyist hardware ... "
**SOURCE:** $1
jage9 2 points 3y ago
very intriguing. Too bad I missed the early-bird, but will definitely think about this. Seems like it may be possible to do more advanced object recognition than what's on a phone, especially with the multi-camera setup.
MRXGray [OP] 2 points 3y ago
My R&D work focuses on real time, simultaneous multi-object detection, tracking, neural classification, semantic segmentation, binocular depth-sensing and distance-estimation, OCR in the wild and OCR for physical documents, packaged in a wearable, low power form factor — And my preliminary iterative tests with OAK-D continues to generate promising results ...
Plus, after brainstorming a few days ago with Peter Meijer who created SeeingWithSound — I've added another target in the mix — Virtual 3D audio.
No need for panning beep sounds — Dynamically generated spoken TTS classifications integrated into a real time immersive 3D soundscape can already contain clockface placement and depth / distance information.
And it's quite straightforward to implement this along with real time, simultaneous multi-object depth-sensing, distance estimation and ML / DL classification methods and algorithms ...
And I rarely use mobile apps for object recognition.
Just mainly SeeingAI's OCR.
But only when I'm sitting, indoors.
I'm not a fan of holding my iPhone and pointing it at various directions I want described while walking around and holding my cane on the other.
Plus, still no depth-sensing and distance estimation ...
Anyway, OAK-D is quite promising indeed for the not-for-profit R&D work that I'm doing.
A year ago, I built open source, 3D printed DIY eyeglasses for blind users through an R&D grant from our national government.
But it's still a little bit on the bulky side. And it's a little heavy for optimum comfort.
And no binocular depth-sensing and distance estimation — Just central view distance estimation through sonar sensors.
Here's a preliminary video that briefly documents my R&D progress: https://www.youtube.com/watch?v=PB9R9DMvgug
Plus, after brainstorming a few days ago with Peter Meijer who created SeeingWithSound — I've added another target in the mix — Virtual 3D audio.
No need for panning beep sounds — Dynamically generated spoken TTS classifications integrated into a real time immersive 3D soundscape can already contain clockface placement and depth / distance information.
And it's quite straightforward to implement this along with real time, simultaneous multi-object depth-sensing, distance estimation and ML / DL classification methods and algorithms ...
And I rarely use mobile apps for object recognition.
Just mainly SeeingAI's OCR.
But only when I'm sitting, indoors.
I'm not a fan of holding my iPhone and pointing it at various directions I want described while walking around and holding my cane on the other.
Plus, still no depth-sensing and distance estimation ...
Anyway, OAK-D is quite promising indeed for the not-for-profit R&D work that I'm doing.
A year ago, I built open source, 3D printed DIY eyeglasses for blind users through an R&D grant from our national government.
But it's still a little bit on the bulky side. And it's a little heavy for optimum comfort.
And no binocular depth-sensing and distance estimation — Just central view distance estimation through sonar sensors.
Here's a preliminary video that briefly documents my R&D progress: https://www.youtube.com/watch?v=PB9R9DMvgug
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.