Cognitive Systems Corp.

Mikhail Ravkine

Oleksiy Kravets

In a general aspect, a motion detection system manages leaf nodes used for sounding by one or more access points. For example, an access point identifies one or more static leaf nodes based on presence activity of each leaf node in a calibration window. A health score is determined for each AP-leaf node link for each calibration window based on AP-leaf node link quality information. One or more of the static leaf nodes are selected to be used for sounding in the motion detection system based on the health scores for each of the AP-leaf node links. The motion detection system is then updated to use the selected one or more static leaf nodes for motion detection.

Classifying static leaf nodes in a motion detection system

In a general aspect, a motion detection system manages leaf nodes used for sounding by one or more access points. For example, an access point obtains presence information for a plurality of AP-leaf node links for a plurality of calibration periods. Presence activity is determined for each AP-leaf node link in each calibration period based on its respective presence information. Static leaf nodes are identified based on the presence activity for the plurality of AP-leaf node links in a calibration window, the calibration window comprising a number of the plurality of calibration periods. The motion detection system is updated to use at least one of the identified static leaf node as a sounding node for motion detection.

Identifying static leaf nodes in a motion detection system

In a general aspect of the examples described, sensor data from a sensor device are used to improve training data for a motion detection system. In some aspects, 'ground truth' labels, sensor data, and channel information are received for a motion detection training period. The labels and sensor data are analyzed to detect variances between the labels and sensor data. The labels are corrected or augmented based on the sensor data. The channel information is tagged with the labels and provided as training information to train the motion detection system.

Growing community of inventors

Waterloo, Canada

Mikhail Ravkine