Cognitive and sensory accessibility to video conferencing systems (ACCESS2MEET)
A significant portion of technologies involves interacting with people through user interfaces in interactive systems. Therefore, it is necessary to design accessible user interfaces which ensure that all people can access and operate these systems regardless of their visual, auditory, cognitive or motor abilities. For this reason, the primary aim of this project is to show the viability and application of accessibility techniques for the design and development of of video conferencing systems that ensure persons with disabilities (PWDs) can access and operate them. Due to the nature of accessibility requirements, this project will use a multidisciplinary approach based on the shared scientific knowledge that methods from fields such as Artificial Intelligence (AI) and Human-Computer Interaction (HCI) can provide.
There are users with sensory impairments, such as blind people, individuals with low vision, deaf individuals, or those with hearing impairments, who need video conferencing applications that include necessary accessibility requirements, such as services providing audio descriptions of videos and images, interfaces that are adapted in relation to sensory characteristics, subtitling services, voice-operated applications or screen readers. Furthermore, there are cognitive accessibility barriers that affect both people with intellectual disabilities and the elderly, all of whom require intuitive user interfaces, text simplification services, and the generation for text summarization summaries, among other tools, to assist in the understanding of texts, as set out in this project.
As specific contributions, design standards will be defined following HCI methods for the design of accessible and adaptable user interfaces. Additionally, affective computing and computational semiotic techniques will be explored with regards to their application in the interactive elements of user interfaces. Moreover, telematic techniques used to generate subtitles will be researched, as well as the use of AI techniques for the generation of high-quality transcriptions through the post-processing of subtitles. In the area of AI and Natural Language Processing (NLP), approaches will be explored to carry out the lexical simplification and the creation of corpora in the fields of Easy-to-Read texts and Plain Language. Furthermore, deep learning techniques used to text summarization will also be researched.