https://www.laprensalatina.com/google-equips-pixel-6-with-its-own...

https://www.laprensalatina.com/google-equips-pixel-6-with-its-own-chip-to-exploit-ai-to-the-max/

San Francisco, Aug 2 (EFE).-

For the first time, Google will outfit its new Pixel 6 smartphone with a chip manufactured by the company and designed specifically to exploit artificial intelligence to the maximum degree possible, above all in the areas of photography and voice recognition.As EFE determined, the new Google phone – the official presentation of which is not scheduled until this fall – offers an extremely high level of quality in terms of image-taking, both still photos and videos, also providing this high quality under low-light conditions.

This significant improvement in image quality has been achieved thanks to the new chip – dubbed the “Tensor” – which uses automatic learning and AI to help polish the photographer’s work and provide, in just milliseconds, a perfect result.Thus, when a user takes a photo, the system immediately “augments” the light captured by two cameras, the main one and another ultra-wide angle one, offering a finalized image that is a mixture of what is viewed through both lenses.

The wide angle image is used to provide the definition, while the main camera is used to correct the colors.Google’s vice president for devices and services, Rick Osterloh, told EFE that the Tensor microprocessor helps to exceed the computation limits that were becoming apparent both in the hardware of other suppliers as well as in Google’s earlier Pixels.

Computational photography has been one of the hallmarks of the Pixel telephones, but this sixth model in the series now achieves new levels of capability.In addition to giving better definition and color, the chip also helps to correct movement in blurry photos and, when decisions need to be made regarding which aspects to improve, it always prioritizes human faces.

Osterloh said that Google’s aim is to provide images that correspond most faithfully to what the human eye perceives.Along with photography, voice recognition is the other area where the Tensor microchip provides significant advances, as – for example – in having the entire process take place within the same device without the need to be connected to the Cloud or even into the Internet.

In this way, when reproducing any video on the phone, if the sound is too low, a user has the option to include subtitles and to translate them into other languages in real time.

One can reproduce, for example, a video made by a friend in English with subtitles in Spanish.

According to Google, the development of the Tensor chip took the firm’s engineers four years of work and the aim right from the start has been to unleash the full potential of artificial intelligence, as that technology stands at present.