In recent years speech recognition and text-to-speech (TTS) conversion has become part of everyday life for many people, whether visually impaired or not. The most familiar technology is provided as interactive audio services, including Amazon's Alexa, Google's Assistant, and Apple's Siri. Many devices, from mobile phones to desktop computers now support speech recognition and synthesis, with a wide range of selectable voices, languages and related settings. In addition, speech-to-text (STT) technologies are having a renewed level of interest and success, despite many earlier attempts to develop this area of technology. Amazon's Transcribe program is a good example of this, as is Google Speech-to-Text cloud service.
To date there have been relatively few concerns about the security of speech enabled devices and associated software, although there are some obvious issues of privacy and biometrics that will arise as such systems become more widely used.
For users with partial visual impairment, severe dyslexia, and a range of other constraints on reading, facilities exist in the Javelin3 PDF readers to provide text to speech. Specialist screen reader software like NVDA and JAWS will only recognize the functionality of Javelin readers (e.g. menus, icons), not the displayed page text. Within Javelin3 readers the recommended approach is to select a block of text in the document or a page and select the text-to-speech key sequence or pop-up option for that section of text or page to be read aloud. Selected text (where available) is best otherwise the facility will try and read the entire page or book from start to finish, and generally this does not work well with many PDFs because of their complex structure. All Javelin3 readers support this functionality (macOS, Windows, Android), although Javelin for iOS/iPadOS does not (at present). On Windows computers this facility is provided in Javelin3 via a right-click operation and selection from the drop-down menu (read selected text or read page). The voice used to read-aloud can be selected from a list if required, for example to select a Spanish voice when reading aloud Spanish text (will require installation of the Spanish voices option via the Windows/Settings/Time & Language facilities). On Mac computers the Javelin3 reader supports the Mac read-aloud facilities. To use this text-to-speech option, open a PDF or secured PDF file (typically a drmz file) and select a paragraph that you want to hear – press the Option key and Esc at the same time, and the text should be read-aloud. The SETTINGS app on Mac computers controls this process, and depending on the macOS version you are running, it might be under a heading “Dictation and Speech” or “Accessibility, Speech section). The Settings app defines the speed of speech, voice used, and key sequence that starts and stops the read-aloud facility, so please check these if you have any problems. On Android devices, touch the page to see the toolbar and choose the three dots icon to select the Read Current Page option from the pop-up menu - see sample screenshot, below).
For users with more severe conditions, such as total blindness or physical disabilities that make reading print or screen-based materials very challenging or impossible, there are a number of additional options available. In general, the source material, in whatever format, will need to be converted to one or more alternative formats. Examples include production of an audio version of a publication, and the conversion of source text to Braille or formats for specialist reader devices. Web-based display using our Webdoxx services can be set up for screen readers, thereby enabling accessibility to a wide audience of people with a range of visual or cognitive disabilities. Ensuring PDFs and web materials have excellent navigation facilities, and a reasonably linear content structure, helps a great deal. Adobe Acrobat provides a useful set of tool for checking and enhancing the accessibility of PDF documents. A useful video on these issues as they affect web page design in particular, see: here.