The corporate introduced that Gemini Nano capabilities are coming to the corporate’s accessibility function, TalkBack. It is a nice instance of an organization utilizing generative AI to open its software program to extra customers.
Gemini Nano is the smallest model of Google’s large-language-model-based platform, designed to be run solely on-device. Meaning it doesn’t require a community connection to run. Right here this system can be used to create aural descriptions of objects for low-vision and blind customers.
Within the above pop-up, TalkBack refers back to the article of clothes as, “An in depth-up of a black and white gingham gown. The gown is brief, with a collar and lengthy sleeves. It’s tied on the waist with a giant bow.”
In response to the corporate, TalkBack customers encounter round 90 or so unlabeled photographs per day. Utilizing LLMs, the system will be capable of supply perception into content material, probably forgoing the necessity for somebody to enter that info manually.
“This replace will assist fill in lacking info,” Android ecosystem president, Sameer Samat, famous, “whether or not it’s extra particulars about what’s in a photograph that household or buddies despatched or the model and reduce of garments when purchasing on-line.”
The machine can be arriving on Android later this yr. Assuming it really works in addition to it does within the demo, this may very well be a recreation changer for blind individuals and people with low imaginative and prescient.
We’re launching an AI e-newsletter! Join right here to start out receiving it in your inboxes on June 5.
