Slashdot Mirror


Google Awarded Face-To-Unlock Patent

An anonymous reader writes "CNet reports that Google was awarded a patent yesterday for logging into a computing device using face recognition (8,261,090). 'In order for the technology to work, Google's patent requires a camera that can identify a person's face. If that face matches a "predetermined identity," then the person is logged into the respective device. If multiple people want to access a computer, the next person would get in front of the camera, and the device's software would automatically transition to the new user's profile. ... Interestingly, Apple last year filed for a patent related to facial recognition similar to what Google is describing in its own service. That technology would recognize a person's face and use that as the authentication needed to access user profiles or other important information.'"

1 of 194 comments (clear)

  1. Re:Good facial recognition by Cacadril · · Score: 5, Informative
    Patents are supposed to disclose enough to enable a person skilled in the art to recreate the invention. But the problems that a person skilled in the art must overcome to recreate this invention, are thousands of times more demanding than coming up with the details of this claim. How can it be obvious to the person skilled in the art how to implement this invention if the standard of "obviousness" is such that the invention itself is not obvious to the person skilled in the art?

    This patent, like most modern, computer-related patents, do not describe, much less patent, the actual solution to the problem. They patent the problem itself.

    Consider, for instance, claim 12 (for increased legibility, I have added some punctuation, numbering, and line breaks):

    A computer program product
    - stored on a non-transitory tangible computer readable medium
    - and comprising instructions that, when executed, cause a computer system to:
    1. receive an image of the first user via a camera operably coupled with the computing device;
    2. determine an identity of the first user based on the received first image;
    3. if the determined identity of the first user matches the first predetermined identity,
    - then, based at least on the identity of the first user matching the first predetermined identity,
    - log the first user in to the computing device;
    4. receive a second image of a second user via the camera operably coupled with the computing device;
    5. determine an identity of the second user based on the received second image;
    6. and if the determined identity of the second user matches the second predetermined identity,
    - then, issue a prompt to confirm that the first user should be logged off of the computing device
    - and that the second user should be logged on to the computing device;
    7. receive a valid confirmation from the first or second user in response to the prompt;
    8. in response to receiving the valid confirmation,
    - log the first user off of the computing device
    - and log the second user in to the computing device.

    --
    There is no substitute for common sense. Especially, no body of rules will do.