We are talking again about accessibility and Google’s technological efforts to reduce the gaps for those who suffer from a certain type of disability. Within the existing accessibility tools, Google has developed several technologies dedicated to making life easier for those who suffer from a certain type of visual impairment or who have virtually no vision.
To make existing images on the web accessible to more people, Google has just extended support for its untagged, artificial intelligence-based web image description technology built as a feature in Chrome to ten new languages..
This technology was developed by the Accessibility teams of Chrome and Google Research, taking into account that there are still a lot of images on the web that don’t have the mandatory label which allows them to be described to visually impaired people using their screen reading devices.
Google recalls that this technology was launched in 2019 in English and extended to five new languages, including Spanish, in 2020.
In this way, so far supporting English, French, German, Hindi, Italian and Spanish, now, at the end of 2021, ten new languages are joining: Croatian, Czech, Dutch, Finnish, Indonesian, Norwegian, Portuguese, Russian, Swedish and Turkish.
Generally speaking, the company recognizes that while the technology has advanced a lot, it still falls short of what a human can describe. and which currently works well on landscape images, although he finds it very difficult to describe other types of images, whether they are sketches, cartoons, memes, or screenshots.
More explicitly by Google:
Automatically generated image descriptions can be incredibly useful and their quality has changed a lot, but it’s important to note that they still can’t caption every image as well as a human. Our system was designed to describe natural images, and it is unlikely to generate a description for other types of images, such as sketches, cartoons, memes, or screenshots.
Visually impaired people must first activate their screen reader via Chrome and then activate the “Get image descriptions from Google” function, either via the context menu when browsing a web page or in the settings of accessibility of Chrome.
Chrome will take care of providing descriptions of images on the web that don’t have labels in their own language.
More information: Google