Google is entering into an official collaboration with Silicon Valley chip maker Movidius, to help realize more effectively audacious image recognition technology straight on smartphones. In devices we have as of today, capabilities in this dimension are basically limited to cloud-based apps, which enact communication with servers to get into advanced artificial intelligence techniques for undertaking tasks in the mould of facial recognition as well as other tasks.
This partnership to say , however, will usher in the reality of Google placing Movidius’ MA2450 chip inside Android handsets so as to give them the capacity to take on tasks like understanding and even identifying images street signs like faces in real time, without the strict requirement to upload photos and patiently cross your fingers waiting for algorithms to perform their dazzling duty in the cloud.
If we infer this largely, the prospect is handsome as individuals with disabilities — an image-recognizing smartphone could possibly contribute invaluably in helping blind individuals more easily get around urban areas or even hold discussions with people they are acquainted on the street. It will be an extension of what Google is at the moment does with Google Translate, which works offline entirely offline allowing users to hold their phone in front of words in a foreign language and then get the text translated on the screen.
In this innovative push, Google strive to unknot the difficult problem which comes with the recognition of objects and people at large, extending this past text. Google Photos, the company’s storage software, already has this service on the cloud, giving you the functionality to search for things like flowers. mountains and see matching photos- but particularly restricted to when connected to Google servers.