Gemini Live gets smarter: Google adds camera recognition and app integration
Google has introduced a series of new features for its virtual assistant, Gemini Live. The main innovations include the assistant's ability to recognize objects through a smartphone's camera, deeper integration with applications, and an improved audio model.
Gemini Live will now be able to recognize objects a user shows it via the camera and provide recommendations based on them. For instance, the assistant could help a user select the correct tool for a job. Additionally, Gemini Live will learn to interact with other applications: if you are discussing a route, the assistant will be able to automatically send a message about your delay using data from the Messages and Contacts apps.
The audio model for Gemini Live has also been updated, allowing it to more accurately recognize a user's intonation and tone of voice. In turn, the assistant itself will learn to vary its own intonation and speech speed to make conversations feel more natural. It will even be able to dramatically retell a story, mimicking a specific character. The new features will first become available on Google Pixel 10 smartphones, with a wider rollout to other devices to follow.
ORIENT
