Intuitive computing methods and systems
First Claim
1. A method practiced by a battery-powered mobile wireless communications device equipped with a camera and microphone, the device having multiple recognition modes enabling recognition of multiple types of content, the method comprising the acts:
- initiating one or more stages of a first recognition agent process to recognize audio or image content data captured by the mobile device, the first recognition agent process performing a recognition selected from a list consisting of;
image watermark recognition, image recognition, object recognition, facial recognition, barcode recognition, optical character recognition, audio watermark recognition, speech recognition, speaker recognition, or music recognition;
receiving detection state data indicating a state of the first recognition agent process in performing the selected recognition;
after initiating the first recognition agent process, receiving first user interest data indicating interest of the user in obtaining a result of the first recognition agent process; and
varying an allocation of processing resources to the first recognition agent process based on both said detection state data and on said user interest data;
the method thereby optimizing use of the device battery by allocating processing resources in response both to the state of the first recognition agent process in performing the selected recognition, and to user input received after the first recognition agent process has been initiated, indicating user interest in the result of the first recognition agent process.
0 Assignments
0 Petitions
Accused Products
Abstract
In one particular aspect, a portable computing device (e.g., a tablet or smartphone) senses audio and/or image content from a user'"'"'s environment, and initiates one or more recognition agents (e.g., performing image watermark recognition, image recognition, object recognition, facial recognition, barcode recognition, optical character recognition, audio watermark recognition, speech recognition, speaker recognition, or music recognition). Resource allocation to a recognition agent can be varied based on (a) progress of the recognition agent to achieve its recognition goal, and (b) user interest data indicating user interest in the output of the recognition agent. A second candidate recognition agent can be evaluated for possible launch, based on a relevance score, and a cost score. In some embodiments, the device adapts its operation to changing context, by terminating a first recognition agent in favor of a second recognition agent, without express user instruction to do so. A great number of other features and arrangements are also detailed.
35 Citations
21 Claims
-
1. A method practiced by a battery-powered mobile wireless communications device equipped with a camera and microphone, the device having multiple recognition modes enabling recognition of multiple types of content, the method comprising the acts:
-
initiating one or more stages of a first recognition agent process to recognize audio or image content data captured by the mobile device, the first recognition agent process performing a recognition selected from a list consisting of;
image watermark recognition, image recognition, object recognition, facial recognition, barcode recognition, optical character recognition, audio watermark recognition, speech recognition, speaker recognition, or music recognition;receiving detection state data indicating a state of the first recognition agent process in performing the selected recognition; after initiating the first recognition agent process, receiving first user interest data indicating interest of the user in obtaining a result of the first recognition agent process; and varying an allocation of processing resources to the first recognition agent process based on both said detection state data and on said user interest data; the method thereby optimizing use of the device battery by allocating processing resources in response both to the state of the first recognition agent process in performing the selected recognition, and to user input received after the first recognition agent process has been initiated, indicating user interest in the result of the first recognition agent process. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)
-
-
16. A method practiced by a battery-powered mobile wireless communications device equipped with a processor, memory, wireless communication interface, camera and microphone, the device having multiple recognition modes enabling recognition of multiple types of content, the method comprising the acts:
-
initiating one or more stages of a first recognition agent process to recognize audio or image content data captured by the mobile device, the first recognition agent process performing a recognition selected from the list consisting of;
image watermark recognition, image recognition, object recognition, facial recognition, barcode recognition, optical character recognition, audio watermark recognition, speech recognition, speaker recognition, or music recognition;receiving detection state data indicating a state of the first recognition agent process in performing the selected recognition; determining a relevance score for a second recognition agent process that the mobile device may run, the relevance score being based on data including at least one data selected from a list consisting of;
location, sensor data available in the memory, context, expressed user intent, and user history;determining a cost score for said second recognition agent process, the cost score being based on data including at least one data selected from a list consisting of;
memory usage, processor usage, and communication bandwidth; andvarying an allocation of processing resources to the first recognition agent process based on at least (a) the detection state data for the first recognition agent process, (b) the relevance score of the second recognition agent process, and (c) the cost score for the second recognition agent process. - View Dependent Claims (17, 18)
-
-
19. A non-transitory computer readable medium containing software instructions adapted to configure a battery-powered, camera- and microphone-equipped mobile wireless communications device to perform acts including:
-
initiating one or more stages of a first recognition agent process to recognize audio or image content data captured by the mobile device, the first recognition agent process performing a recognition selected from a list consisting of;
image watermark recognition, image recognition, object recognition, facial recognition, barcode recognition, optical character recognition, audio watermark recognition, speech recognition, speaker recognition, or music recognition;receiving detection state data indicating a state of the first recognition agent process in performing the selected recognition; after initiating the first recognition agent process, receiving first user interest data indicating interest of the user in obtaining a result of the first recognition agent process; and varying an allocation of processing resources to the first recognition agent process based on both said detection state data and on said user interest data; wherein said software instructions serve to optimize use of the device battery by allocating processing resources in response both to the state of the first recognition agent process in performing the selected recognition, and to user input received after the first recognition agent process has been initiated, indicating user interest in the result of the first recognition agent process. - View Dependent Claims (20)
-
-
21. A battery-powered wireless system including:
-
plural sensors, including a camera and a microphone; a wireless communications interface; a battery; one or more battery-powered processors; and a memory including software instructions, the software instructions configuring the system to perform acts including; initiating one or more stages of a first recognition agent process to recognize audio or image content data captured by one of said sensors, the first recognition agent process performing a recognition selected from a list consisting of;
image watermark recognition, image recognition, object recognition, facial recognition, barcode recognition, optical character recognition, audio watermark recognition, speech recognition, speaker recognition, or music recognition;receiving detection state data indicating a state of the first recognition agent process in performing the selected recognition; after initiating the first recognition agent process, receiving first user interest data indicating interest of the user in obtaining a result of the first recognition agent process; and varying an allocation of processing resources to the first recognition agent process based on both said detection state data and on said user interest data; wherein the system is configured to optimize use of the battery by allocating processing resources in response both to the state of the first recognition agent process in performing the selected recognition, and to user input received after the first recognition agent process has been initiated, indicating user interest in the result of the first recognition agent process.
-
Specification