-
公开(公告)号:US20220386260A1
公开(公告)日:2022-12-01
申请号:US17886068
申请日:2022-08-11
Applicant: Microsoft Technology Licensing, LLC
Inventor: Woo Suk LEE , Flavio Protasio RIBEIRO , Alexander POPOVICH
IPC: H04W56/00
Abstract: Individual clock adjustments between electronic devices are typically based around a round-trip time (RTT) measurement of the reference message between initiating and the receiving devices. With increasing expectations of clock synchronization accuracy, as well as widespread use of wireless data networks, the presently disclosed technology provides a dedicated clock synchronization network that yields a fixed delay between hops and within associated devices of a dedicated clock synchronization network. By accounting for the known delays between hops and within associated devices of the dedicated clock synchronization network, better clock synchronization accuracy can be achieved than prior art techniques that estimate latency based on an RTT measurement.
-
公开(公告)号:US20210337498A1
公开(公告)日:2021-10-28
申请号:US17007461
申请日:2020-08-31
Applicant: Microsoft Technology Licensing, LLC
Inventor: Woo Suk LEE , Flavio Protasio RIBEIRO , Alexander POPOVICH
IPC: H04W56/00
Abstract: Individual clock adjustments between electronic devices are typically based around a round-trip time (RTT) measurement of the reference message between initiating and the receiving devices. With increasing expectations of clock synchronization accuracy, as well as widespread use of wireless data networks, the presently disclosed technology provides a dedicated clock synchronization network that yields a fixed delay between hops and within associated devices of a dedicated clock synchronization network. By accounting for the known delays between hops and within associated devices of the dedicated clock synchronization network, better clock synchronization accuracy can be achieved than prior art techniques that estimate latency based on an RTT measurement.
-
公开(公告)号:US20220408013A1
公开(公告)日:2022-12-22
申请号:US17354675
申请日:2021-06-22
Applicant: Microsoft Technology Licensing, LLC
Inventor: Alexander POPOVICH , James Alan HOLT
Abstract: Systems and methods directed to adjusting an image based on a detected object depicted in the image are described. The method may include receiving an image from an image sensor, receiving statistical information associated with the image, detecting an object depicted in the image using a deep neural network, identifying object-specific statistical information for the detected object, generating a weighted object-specific parameter based on the object-specific statistical information, generating a weighted-image value based on the weighted object-specific parameter, providing the weighted-image value to the image sensor, where the image sensor is configured to update one or more image sensor parameters based on the weighted-image value, and acquiring an image from the image sensor updated with the one or more image sensor parameters.
-
-