The number of services supported by generative artificial intelligence and their impact on the labour market and humanity is growing exponentially every day. This article provides a review of solutions that facilitate access to digital culture for those with limited access. Artificial intelligence is a skill-enhancing technology that not only enables those who are good at something to be even better, but can also help overcome disadvantages.
Lance Carr is a famous figure in the gaming world. He has limited mobility and cannot control a computer with his hands, so he used his computer with a head-mounted mouse and became a famous gaming streamer. In 2021, his house burned down. This tragedy prompted the Gameface project, a facial recognition system that uses artificial intelligence to learn and allow its user to control his mouse using facial movements. Users can issue commands by moving their eyebrows, opening and closing their mouths. Head movements are recognized using a simple webcam, and head and facial movements captured by the webcam are recognized by machine learning models that also move the cursor. Users can customize gestures, and they can set the extent of grimaces and face movements that activate certain commands. The software is based on the open source solutions of Google MediaPipe, it can be downloaded for free, and professionals can also contribute to the development. Although the project was prompted by a gamer, it is easy to see the potential for later expansion among those in need.
Smartphones can be used more and more widely to overcome various disadvantages: for example, they can help with navigation, read captions out loud, and run various e-health applications. There are various solutions that are currently used in the field of accessibility, and the generative ability of artificial intelligence makes them even more flexible and usable in all situations. For example, the Ava application transcribes nearby conversations for the hearing impaired, RogerVoice does the same with phone conversations, and Subtitle Viewer connects to the home TV and provides subtitles on the user’s mobile phone.
For example, a person with a mental disability can use text summaries produced by artificial intelligence systems to better understand the world around him. A message that may at first seem like a complicated riddle can be turned into an easy-to-understand text. Things that were difficult or impossible for such persons can become easily accessible on a daily basis. Technology adapts and helps to make the world a more inclusive place with the accessibility of artificial intelligence.
Artificial intelligence already performs very convincingly on the field of accessibility in the following areas:
- automated image recognition and interpretation for the visually impaired
- facial recognition for the visually impaired (not only in a digital environment)
- lip reading for the hearing impaired
- text summaries for people with mental disabilities
- automated real-time subtitling for the hearing impaired and translation for persons who do not understand the given language
For the visually impaired, artificial intelligence-supported systems will become even more robust in the following areas:
- Voice based user interface: AI technologies enable the development of voice-based user interfaces that allow the visually impaired to use voice commands to control a computer. This can help with browsing, launching applications, and retrieving information.
- Voice assistants: Intelligent voice assistants such as Siri, Alexa or Google Assistant allow the visually impaired to give voice instructions for tasks to be performed on the computer. These assistants can search for information, send messages, manage calendar events, and perform other activities for the visually impaired.
- Image recognition and object recognition: AI algorithms use image recognition and object recognition systems to help the visually impaired identify and name objects or people around them. This technology allows, for example, reading materials such as books, newspapers or spreadsheets to be scanned and voiced.
- Voice navigation: AI-based voice navigation systems can help visually impaired people navigate computers and switch between applications. For example, screen readers are software that read on-screen text and allow blind or visually impaired users to understand the content of the text.
For the hearing impaired, text recognition and speech generation enables users to read and understand conversations, podcasts, videos, or other audio content in text form. In addition, AI can generate speech and convert written text back into speech, meaning that text content becomes accessible to hearing-impaired users as well. With the help of large and possibly online databases, it is now possible for a hearing-impaired person using public transport to receive written messages about important noises or even information broadcast via a public announcement system.
Different solutions already exist and artificial intelligence expands the possibilities by becoming more and more able to solve not only pre-defined situations, but also functioning better and better in life-like, natural, and unscripted environments.
Of course, one should not forget about the legal challenges raised by such applications. They collect extensive data about people in the environment of users, including images of faces, objects in a building, or the contents of written documents. Their extensive use probably requires a clearly visible signal to the environment that systematic data collection is taking place, for the purpose of helping a fellow human being. Just as we cannot ignore that developments that facilitate access and eliminate disadvantages go hand in hand with possibilities for abuse; for example, the recognition and interpretation of images can be used to evade captcha protections, and face scans can be used for deepfake manipulations or even to hack the protection of smartphones.
