ROLE OF AI FOR DISABLED PEOPLE
Technology in general and artificial intelligence in particular play a key role in accessibility. It is not only about finding the latest innovations, but primarily about providing solutions that improve the quality of life by serving certain categories of people. AI has a huge impact on the daily lives of people with disabilities. People with mental disabilities can easily learn about the world around them through text abstraction. What initially seemed like a difficult-to-decipher message turned out to be an easy-to-understand text.
Artificial intelligence technologies such as AI2 surveillance are improving their ability to recognize objects and faces. This application is created by blind developers, talented people with deep knowledge and understanding of the needs of the visually impaired are developing these useful skills, but most technology developers do not have or understand such in-depth knowledge or understanding. You learn about disability and accessibility in college courses.
There is a culture that technology is helping to create and it’s hugely empowering but I want to get really practical now and talk about the technology and do some demonstrations of these applications in earnest in reality now with any demo it may or may not go well we’ll see okay I’ve taken a cheese harness off by the way so it can be a bit more relaxed let’s see if it works great so on the first slide seeing AI.
Microsoft has put in the effort have taken the trouble to make it into something that is a perfect combination of those mainstream applications or technological applications into an app that is going to be hugely impactful is hugely impactful for people with different disabilities so I’m going to plug in this audio they’re not a lot of cables. If people with disabilities are not represented in the data set, the model created is not comprehensive and will not work for everyone.
When this happens, it’s called a “data desert.” Microsoft’s AI for Accessibility program is one way to address the data desert to improve the expression and diversity of people with disabilities in AI. For example, one of our beneficiaries, the University of Texas at Austin, is working to develop the VizWiz dataset, an image dataset that is clearly presented by people who are blind or visually impaired. I believe accessibility is raising the bar for what AI has to offer.
A continuous and holistic approach to design that includes people with disabilities is an important part of the journey. As an example of what happens when you’re not working with people with disabilities, sign language gloves may sound nice and cool, but development didn’t take into account the fact that sign language is more than a hand gesture. Gestures, facial expressions, and unique grammar. If deaf people are involved, more authentic and functional artificial intelligence tools can be developed. It is a good idea to ask three Design Justice questions as you work to understand the invalidity of disability data in your own work.
· Who benefits?
· Who is potentially harmed?
· And who participates?
Systems skills are not limited to physical disabilities. No Isolation, a Norwegian medical technology startup, is implementing a supplement to help combat social isolation and loneliness. According to the No Isolation study, the two most vulnerable groups are children with chronic conditions and children over 80 years of age.
No Isolation addresses this vulnerability for young children by deploying AV1 robots to help people who are physically unable to attend school. Robots use NLP artificial intelligence and machine vision systems to attend surrogate classes that allow children to interact with teachers and other children. One solution can possibly be to focus design efforts on the equal representation of people with disabilities in education data. It can be suggested that a fair representation of persons with disabilities is equivalent to a comprehensive view that fully and accurately classifies all types of disabilities. Artificial intelligence breakthroughs are clearly assisting firms in meeting their legal duties and creating a more diverse workforce. However, there is a need to build employee and employer trust in AI and make it simple to use. The national AI plan should include possible revisions to the Equal Employment Act to enable wider usage of the technology.
