Google utilizes artificial intelligence to enhance accessibility benefits, and the company has announced that the Gemini Nano technology will be integrated into the “TalkBack” accessibility feature, allowing users with special needs to have a clear description of images.
This is an example of a company using advanced technology to expand the usage of its software to a wider range of users.
Gemini Nano has multimedia support, which means it has the ability to process text more effectively while considering other aspects such as images, sounds, and speech.
Gemini Nano does not require an internet connection to function as Google has designed it to work entirely on the device, using Gemini Nano to provide an audio description of objects for people with disabilities.
The TalkBack feature reads out the image description loudly, whether the image is captured through the camera or from the internet. Gemini Nano offers diverse support for multimedia to provide a detailed interpretation of the image.
Google presented an example showcasing a dress and added the TalkBack feature describing it as “a cotton dress in black and white, short with a collar and long sleeves, tied at the waist with a large bow, which serves as a description of the image”.
According to the company, TalkBack users encounter about 90 uncategorized images daily. When using large language models, the feature becomes capable of analyzing content accurately, potentially avoiding the need to manually add that information.
Google stated: “This update will fill in the missing information, whether it is additional details about what is displayed in the image sent by family or friends, or the clothing style and story when shopping online”.
Gemini Nano will introduce the TalkBack feature later this year, which could bring significant changes to the lives of individuals with disabilities.
The Google TalkBack app was initially launched in 2009, where it can read out on-screen content loudly, allowing users to navigate their devices using specific gestures. It also supports voice commands and the default Braille keyboard.
