Ga direct naar inhoud

Will AI give disabled their online ears and eyes?

Sem Enzerink
2019-09-16

Recent legislation, technical guidelines and online redesigns. Many measures are taken to improve the accessibility of websites and mobile apps to disabled. Despite these efforts plenty users still face unreadable pictures and insufficient color contrast. Can these technological deficiencies be solved with yet another technological solution: AI?

The promised

Reading this online article is not as easy as it sounds. Online content may challenge people with visual, auditory, cognitive, speech or physical difficulties. Disabilities often relate to different use of the human senses. Certain AI domains could potentially address these senses. Whenever speech recognition is able to translate talks into text, deaf people should be able to ‘hear’. Whenever image recognition can translate images into audio descriptions, the blind and partially-sighted should be able to ‘see’. Are AI solutions profound enough to complement or replace human senses, giving disabled a ticket to the digital world?

The delivered

AI already meets some accessibility needs and expectations. The AVA app offers for instance live subtitles to conversations. Speech recognition enables the deaf to read spoken communication and understand web videos without subtitles. Google’s Euphonia Project records the voices of people with speech disorders to improve the voice command options of apps that generally understand Oxford English only. Moreover, Seeing AI from Microsoft narrates whatever your camera or smartphone screen captures. It describes for example images in emails, Twitter and WhatsApp, so visually challenged people know what’s going on.

The untapped

AI solutions have created some ‘readable sounds’ and ‘hearable views’ making the special needs of some website and app users no longer so special. Can we expect more solutions? Adding up our aging population, urge for innovation and the global assistive technology market generating USD 30.82 billion by 2024, seems like an opportune formula for developing new assistive AI solutions. Screen software that automatically changes color contrasts, or online forms that recognize sign language may be just around the corner. In this way, AI empowers all of us to be part of our digital societies.

About the author

Sem Enzerink is a Senior Consultant at Capgemini Invent. Building upon his experience in eGovernment and knowledge of societally relevant digital transformations, he supported the European Commission to implement the Web Accessibility Directive across the European Union Member States.