Cataracts, glaucoma, macular degeneration and many other complaints can reduce vision in later life, though age-related reduced range of focus is the most common impairment. Total blindness is rare and it is usual to described visual impairment by two numbers. 6/60 means the person has to approach to 6 metres to see what a normally-sighed person can see at 60 metres. There is workaround using touch, invented by a Frenchman who devised a script using a pattern of six dots – Braille. A new script in later life would be hard, so the human voice and computers provide the best methods.
There are now computer devices that can recognise and synthesise the human voice. To use such a device if your vision is deteriorating, you need to become very fluent with several different modules while you can still see the control and text.
A voice recognition system is one module. Smartphones now come with a proprietary app such as Siri on Apple products, Cortana on Windows, or Bixby on Android. The main charity for blind people in Britain recommends Alexa. They have a large library of talking books, which can be summoned by voice command to Alexa. Amazon Echo SNL on Youtube gives amusing anecdotes about the “pragmatic” ability (opposite of “semantic”) of the software – its capacity to infer the speaker’s intention – despite poor choice of words.
The intelligence is now good enough to cope with alteration in one user’s voice through respiratory infections, alcohol, fatigue, or urgency, and the differences in pitch between man, woman and child voices. Simple syntax will reduce recognition errors: use one clause, subject-verb-object word order and avoidance of words with multiple meanings. An appointments calendar, an alarm for a specific day and time, a timer for a soft-boiled egg and many other executive actions are possible. More interesting types of language such as poetry will need an AI!
An optical character recognition system (OCR) is the second module. As far back as 1914 Emmanuel Goldberg made a device that recognised characters in a single typeface and turned them into telegraph code, with blind people in mind. Modern OCRs can cope with many fonts, upper and lower case and even handwriting. Pleco or Google Translate let you draw a Chinese character with your fingertip, and present you with the nearest standard typeface characters to choose from. A secretary with failing eyesight can set up an Ipad or other tablet on a trestle and slide a typewritten letter underneath. The software can then recognise text and submit to a speech synthesiser with your preferred accent. Practice involves defining the boundary carefully, to exclude stray text being pronounced and confusing the listener.
A Translation Engine is the third module. The input is normally text in one language and the output is text in the script of the target language. Google Translate, Microsoft Translator, DeepL, Amazon Translate, Apertium and Babelfish are widely used. The main issue for visually-impaired people is likely to be the multiplicity of adaptations made for them, by local libraries, lifts, and bespoke devices. Braille was devised for the French alphabet, so has accented vowels not needed in English, but this creates few problems. Braille can in principle represent text in many languages, but Chinese has to be transliterated into Roman letters, with tones omitted, so ambiguities arise.
A person losing sight in later years is unlikely to learn braille, and will infrequently need to translate from other languages, but will need to be versatile between the various adaptations competing for their attention.
In your smartphone select your translation engine, such as Google Lens. Find the camera icon 📷. (This is unicode character “U+1F4F7”). Move the phone in or out to exclude text in the surroundings leaving only the script below with a large white margin. Select speech output. You do no need to know language it is, as the software will detect it. You should be able to her speech output in English which continues this theme.
大家好，欢迎第三纪大学的各位同学。
在本课程中，我们将练习解决视力障碍的方法。 有一种由法国人发明的众所周知的方法，其中触觉使用六个点的图案。 这对于晚年视力下降的人来说是很困难的，所以人声是最好的方法。
现在已经有了可以识别和合成人声的计算机设备。 英国主要的盲人慈善机构推荐了 Alexa。 他们有一个很大的有声书籍图书馆。 可以通过 Alexa 的语音命令来召唤一本有声读物。
约会日历非常有用。 您可以为特定的日期和时间设置闹钟，它会在时间临近时告诉您。 您还可以设置一个较短的计时器，例如“Alexa，告诉我我的半熟鸡蛋已经过了五分钟”。
可以用外语进行对话，尽管还没有达到同声翻译的水平。 您的翻译单词需要具有精确的语法。 例如，“Alexa，我用南非荷兰语怎么说‘恭喜达到 B 级’？
您将练习为盲人提供行走方向。 你的参考必须是绝对的而不是相对的，所以“它在另一边”和“你在最后一个拐角处走错了”太含糊了。 你需要对他们的词汇做出最好的猜测。 他们能理解这些短语吗？ “右转90度”。 “你正在向东走”。 “路径有 18 英寸宽”。
再见， 埃德
Practice with speech recognition, OCR and translation engines on different apps in different tablets and phones is a good way to work around deteriorating vision.