Online education has become a reality in many universities worldwide, either as complementary educational material in support for conventional courses or as a standalone online information. The web can also offer great support to disabled people who use it for purposes of entertainment, education and communication. With standards emerging in web accessibility, developers are looking into ways to make online material as accessible as possible. Within this context, Audiovisual Translation (AVT) and Assistive Technology (AST) have the potential to create new channels in the provision of university-level educational material on the web for people with sensory impairments such as blindness and deafness.

Digital literacy and accessibility, in a broad sense, is included in the availability of products, services and places to people with all sorts of disabilities – physical, cognitive, mental, sensory, emotional, and developmental – as mandated by the European Union. Accessible education on the web is composed of two major aspects: (a) accessible audiovisual material (i.e. videos) and (b) accessible functional context (i.e. design, navigation, assistive tools, etc.). Accessibility in online education can be achieved through access services and assistive tools, which are respectively the outcomes of Audiovisual Translation (AVT) and Assistive Technology (AST).

Assistive technology such as audio and video captioning and crowd services integrated with e-learning software can help sight and hearing impaired people in education.

Audiovisual Translation and Assistive Technology

AVT services consist of software, such as subtitling for deaf and hard-of-hearing (SDH) and audio description (AD) for blind and partially sighted audiences. AST tools, such as speech recognition, screen reading and magnification of content, are implemented in websites and online platforms, but are also used offline to assist students with several applications. SDH and AD are two services that were born within the field of AVT and today they are established access services used mainly in television and theatrical performances.

SDH, in particular, includes audiovisual material broadcasted or watched in any form of distribution, and can be both intra-lingual (performed within the same language) and inter-lingual (between one or more different languages). AD is a process that "provides a narration of the visual elements – action, costumes, settings – in theatre, television/film, museums exhibitions, and other events" and it "allows patrons who are blind or have low-vision the opportunity to experience arts events more completely – the visual made verbal"1. SDH is greatly valued as a service that advances learners’ reading and writing skills in the same way that AD enhances speaking and listening skills.

Within the educational sphere, the two services have been mainly focused on facilitating learning for students with sensory impairments, but they have been also used for learning languages. These technologies might also be used to increase ease of use for online interaction, and for accessibility of applications such as video games.

Improvements in web accessibility have played a crucial role in shifting from "Web for some" to a more democratic "Web for all"

Within AST, for web and computer-based educational content, available tools consist mainly of software and hardware that make services and content available to disabled people on specific devices. Software developments, such as screen readers, are gradually substituting hardware since they have proved to be affordable, more practical and multifunctional. People with sensory impairments use assistive technologies in their everyday life through commonly used devices, including iPads and smartphones, while several applications are implemented by specific hosts or providers to make life easier for people visiting online websites.

Creating a Web for All

The web is gradually evolving into a sphere with vast potential for people with disabilities. As mandated by regulatory bodies such as the European Union, and via Web Content Accessibility Guidelines (WCAG) of the World Wide Web Consortium (W3C), alongside navigation options for blind and partially sighted users, developers are often required to provide alternative tools for audiovisual material on their websites. This can take the form of pre-recorded audio-only and video-only media, captions, audio description for pre-recorded media, as well as live captions for live audio content in synchronised media.

Initially SDH and AD were mostly available on the web either in the form of amateur services provided by AVT enthusiasts, or as a component of pre-recorded material published on the web. Today, website development language HTML5 facilitates developers in including other forms of SDH and AD. For example, it supports direct embedding of supporting material for multi-media content, e.g. video content, without the need for third party plug-ins. It is safe to say that this development has allowed AVT to enter the world of online accessibility which was until recently restricted only to AST applications, with providers focusing on alternative texts for images and readable content. These improvements in web accessibility have played a crucial role in shifting from "Web for some" to a more democratic "Web for all".

Nevertheless, some web developers have been generating products for online education which do not integrate functional and contextual developments from one field to the other, therefore limiting the benefits to specific disabilities. Ideally the accessibility to AVT and AST technologies should be more flexible and global, and should encompass technical, social, educational and linguistic dimensions.

Crowd-Sourced Solutions?

New AVT services have been gradually used by several institutions for educational purposes. For example, in the field of synthesized speech and voice recognition, new tools such as Synthesized AD, speech-to-text narration, video description and annotation, attempt to substitute audience-targeted and humanly produced SDH and AD. Subtitling and crowd-sourcing have also lately improved their accessibility through more sophisticated means. Open source software and platforms hosting audiovisual material, which includes MAGpie, YouTube and Amara (Universal Subtitles) offer users a simple working environment in order to create captions for their videos or to acquire captions that have been produced through machine translation and speech recognition.

Access services are beginning to flourish both online and offline within the academic environment

At the same time, networks of crowd-sourced and cloud-based subtitling are being built in an attempt to produce fast real-time captioning with minimum latency and maximum precision with hybrid contributors, rather than just humans. An indicative example of crowd-sourced subtitling is Legion:Scribe, a system that, as claimed by the company, "captions speech on-demand and with less than 3 seconds latency by automatically merging the simultaneous input of multiple crowd workers".

A number of products in the field of AST have also emerged to provide more inclusive solutions for online education. Most of them, including ClassInFocus (2009), DELE (2012), SSTAT (2012), MVP (2011) and the Photonote system (2007) focus on assisting either deaf and hard-of-hearing or blind and partially sighted students. Most of the research in AST conducted by universities worldwide focuses on captioning for deaf students. Among the most prominent commercial solutions are systems such as Panopto, Tegrity, MediaSite and Echo306, which form learning environments that capture video, audio and screen activity. They support captions, whether these are produced by people or machines (speech-to-text technologies), and offer users the opportunity to edit videos, make video annotations and provide access to further educational material i.e. PowerPoint presentations. Other open-source solutions for AD and captioning include Amara (captioning), YouTube (captioning), MAGpie (AD), CapScribe (AD) and LiveDescribe (AD). These are proper tools rather than learning solutions, which can satisfy the needs of both blind and deaf students.

AVT and AST in the Classroom

An increasing number of universities have Disability Advisory Centres and Accessibility Units, which offer advice and solutions to disabled students in order to minimize the hurdles to be overcome to attend courses. At Imperial College, besides the well-established Advisory Centre, a new organization – the Assistive Technology Steering Group – was set up in 2012 with the aim of examining issues and providing solutions for access to educational materials by disabled students.

Networks of crowd-sourced and cloud-based subtitling are being built in an attempt to produce fast real-time captioning with minimum latency and maximum precision

The main purpose of the organization is to assist departments in developing new ideas for accessing educational material on the IC website and on on-site computers. The team is now looking into ways to provide lectures online as well as solutions for a more accessible IC website. Moreover, a new European project called ClipFlair (2011) aims to develop an online social network to provide material for learning languages through a series of access services (including captioning and re-voicing) and lesson plans which allow learners to practise their speaking, listening, reading and writing skills.

An important limit of all the aforementioned tools is the lack of holistic solutions for all students with sensory impairments. In most of the cases, users are provided with rather limited applications, i.e. video input, and also when lectures/videos are accompanied by other material, such as presentations, these are only intended for the deaf and hard-of-hearing. Moreover, AVT and AST tools and services are rarely combined in educational contexts, i.e., a screen reading tool smoothly connected to an AD time-coded script.

Another urgent issue is that of quality of access services. Crowd-sourced or transcript captions are not as efficient as SDH. The latter is produced to serve deaf and hard-of-hearing audiences through the use of several techniques such as the preparation of a new script read by a human voice instead of speech synthesis. More importantly, what is missing is a functional approach that allows students to enter learning environments as self-serviced individuals, an idea introduced by Cornford and Pollock in 20032. Other solutions should be critically evaluated in place of inappropriate educational tools such as synthetic and machine-oriented solutions. The present guidelines may often make demands for accessible AV material, but still do not specify the production process and do not openly explain which approaches should be followed by educational institutions in using the relevant material. AST tools and access services should be combined in such a way that they can bridge the gap between the several disciplines.

It is clear that access services are beginning to flourish both online and offline within the academic environment. The key action to be taken in order to guarantee the widest and most efficient accessibility in online education is through the creation of a holistic accessible environment for disabled students incorporating and integrating both AST and AVT technology. This approach will guarantee several benefits such as flexibility and compatibility for different countries while offering social inclusion and interaction throughout the web.