Abstract: A system may be configured to generate real-time analytics using machine learning and computer vision. In some aspects, the system may receive a first video frame from a first video capture device positioned to capture activity at a location within a monitored area, receive a second video frame from a second video capture device positioned to capture interaction activity of customers, determine a first inference based on the first video frame, determine a second inference based on the second video frame, and associate the first inference and the second inference based on determining that the first inference and the second inference correspond to a common time period and common location.

Abstract: Disclosed herein are apparatuses and methods for auto-configuring a region of interest (ROI) associated with a camera. In one implementation, A method includes tracking, over a plurality of image frames spanning a first period of time, a plurality of objects moving in an environment, and determining a plurality of respective trajectories of the plurality of objects. The method includes identifying a common trajectory shared by more than one object of the plurality of objects based on comparing each of the plurality of respective trajectories to one another and identifying an ROI that encompasses the common trajectory. The method further includes tracking, over a plurality of new image frames spanning a second period of time, another plurality of objects moving in the environment, and generating an alert that the common trajectory has changed based on new trajectories of the another plurality of objects.

Abstract: Aspects of the present disclosure relate generally to a vision system that detects objects moving in an environment. The vision system detects objects that crossed a boundary of an environment during a first period of time, and determines a count of objects that crossed the boundary in a first direction. The vision system retrieves historical egress and ingress data for a second period of time corresponding to the first period of time. The historical egress and ingress data comprises a detected historic count of crossings in the first direction and a corrected historic count of crossings in the first direction. The vision system calculates an error rate for the second period of time based on a ratio of the detected historic count and the corrected historic count, and determines a corrected count for the first period of time by adjusting the count using the error rate.

Abstract: Systems and methods for object threat detection are discussed. Techniques for performing a hybrid analysis threat assessment using both a first artificial intelligence model trained on known threats and a second artificial intelligence model trained to detect anomalies are described.

Abstract: The present disclosure relates to a system for determining anomalous behavior. The system comprises a processing device that is configured to generate first segmentation information and second segmentation information from the video feed, the first segmentation information corresponding to a first potential occupant and the second segmentation information corresponding to a second potential occupant. The processing device further configured to determine that the first segmentation information and the second segmentation information correspond to actual occupants based on the first segmentation information and the second segmentation information, and generate a first bounding box associated with the first segmentation information and a second bounding box associated with the second segmentation information.

Abstract: A system may be configured to generate real-time analytics using machine learning and computer vision. In some aspects, the system may receive a storage structure video frame and an interaction video frame, determine a storage structure inference based on the storage structure video frame, and determine an interaction inference based on the interaction video frame. Further, the system may determine that the storage structure inference and the interaction inference correspond to a common time period and common location, and generate analytics information based on the storage structure inference and the interaction inference.

Abstract: Disclosed herein are apparatuses and methods for tracking occupancy count in an environment based on detected movements and dwell time. An implementation may comprise detecting a person in a plurality of image frames captured by a camera. The implementation may comprise tracking a movement of the person and determining that the person has crossed a boundary within the plurality of image frames. The implementation may comprise identifying a path of the person based on the tracked movement and may comprise calculating a dwell time in response to determining that a difference between the identified path and at least one target path associated with entering or exiting the environment exceeds a threshold difference. The implementation may comprise comparing the dwell time to a target dwell time, and updating and storing the occupancy count of the environment in response to determining that the dwell time is less than the target dwell time.

Abstract: Example aspects include a method, apparatus and computer-readable medium of determining losses at a point of sale (POS) device, comprising receiving, by a processor from an imaging device, a video feed of a scanning area. The aspects further include detecting, by the processor, an entry of an item into the scanning area. Additionally, the aspects further include identifying, by the processor, one or more motion parameters of the item. Additionally, the aspects further include determining, by the processor, a dwell-time for the item based at least on the one or more motion parameters. Additionally, the aspects further include identifying, by the processor, a scan time anomaly for the item. Additionally, the aspects further include outputting a notification, by the processor, indicating a suspicious activity for the item, wherein the notification indicating the suspicious activity is based on the scan time anomaly.

Abstract: Disclosed herein are apparatuses and methods for iteratively mapping a layout of an environment. The implementations include receiving a visual stream from a camera installed in the environment, wherein the visual stream depicts a view of the environment, and wherein positional parameters of the camera and dimensions of the environment are set to arbitrary values. The implementations include monitoring a plurality of persons in the visual stream. For each person in the plurality of persons, the implementations further includes identifying a respective path that the person moves along in the view, updating the dimensions of the environment captured in the view, based on an estimated height of the person and movement speed along the respective path, and updating the positional parameters of the camera based on the updated dimensions of the environment. The implementations further includes mapping a layout of the environment captured in the view of the camera.

Abstract: Example implementations include a method, apparatus and computer-readable medium for person detection by a computer device, including retrieving, from a database of suspicious persons, a profile of a person detected in at least one image, wherein the profile lists crime incidents associated with the person. The implementations include determining, using the profile, a scale of crimes committed by the person based on an amount of time between incidents of a certain type, wherein the certain type is one of theft at a store or theft of a particular product type. The implementations include determining a rank of the person in the database of suspicious persons based on a frequency and the scale of crimes committed by the person compared to other suspicious persons, and transmitting, to a second computer device, the retrieved profile and a notification that indicates that the person is detected when the rank exceeds a threshold rank.

Abstract: Aspects of the present disclosure relate generally to a vision system that detects objects moving in an environment. The vision system detects objects that crossed a boundary of an environment during a first period of time, and determines a count of objects that crossed the boundary in a first direction. The vision system retrieves historical egress and ingress data for a second period of time corresponding to the first period of time. The historical egress and ingress data comprises a detected historic count of crossings in the first direction and a corrected historic count of crossings in the first direction. The vision system calculates an error rate for the second period of time based on a ratio of the detected historic count and the corrected historic count, and determines a corrected count for the first period of time by adjusting the count using the error rate.

Abstract: Aspects of the present disclosure include methods, systems, and non-transitory computer readable media for receiving an initial image of the area, receiving an updated image of the area, detecting at least one of a person in the area or a movement in the area based on the updated image, comparing the updated image and the initial image to identify a change in the area, and determining a presence or an absence of the occlusion based on the change in the area and the at least one of the person or the movement in the area.

Abstract: Example implementations include a method, apparatus and computer-readable medium for person detection by a computer device, comprising receiving a plurality of video frames from a camera located in an environment, wherein the video frames include a depiction of a person in the environment. The implementations further include identifying attributes of the person from the video frames and comparing the attributes with attribute entries in a database of suspicious persons, wherein the database of suspicious persons includes attributes of a plurality of persons associated with an alert in the environment. The implementations include retrieving a profile indicative of an alert history of the suspicious person in response to determining a match between the attributes of the person and an entry in the database corresponding to a suspicious person, and transmitting, to a second computer device, the retrieved profile and a notification that indicates that the suspicious person is in the environment.

Abstract: Aspects of the present disclosure relate generally to a vision system that detects persons exiting an environment, comprising detecting, by a processor using a plurality of image frames from at least one sensor, persons that exited the environment during a first period of time. The vision system further determines, by the processor, an exit count for the first period of time and retrieves, from a database, historical egress data comprising a detected historic exit count and a corrected historic exit count of the environment for a second period of time corresponding to the first period of time. The vision system calculates an error rate for the second period of time based on a ratio of the detected historic exit count and the corrected historic exit count, determines and stores a corrected exit count for the first period of time by adjusting the exit count using the error rate.

Abstract: Disclosed herein are apparatuses and methods for tracking occupancy count in an environment based on detected movements and dwell time. An implementation may comprise detecting a person in a plurality of image frames captured by a camera. The implementation may comprise tracking a movement of the person and determining that the person has crossed a boundary within the plurality of image frames. The implementation may comprise identifying a path of the person based on the tracked movement and may comprise calculating a dwell time in response to determining that a difference between the identified path and at least one target path associated with entering or exiting the environment exceeds a threshold difference. The implementation may comprise comparing the dwell time to a target dwell time, and updating and storing the occupancy count of the environment in response to determining that the dwell time is less than the target dwell time.

Abstract: Disclosed herein are apparatuses and methods for providing contextual data for a security event in an environment. An implementation may comprise receiving and parsing sensor data from a plurality of sensors located in the environment. The implementation may comprise storing, in a database, the parsed sensor data comprising identifiers of the plurality of objects and the attributes. The implementation may comprise detecting a security event at the environment and determining a type, a time window, and a location of the security event. The implementation may comprise retrieving, from the database, a set of object identifiers with timestamps within the time window and a matching location of the security event. The implementation may comprise filtering, from the set of object identifiers, at least one object identifier that matches the type of the security event, and outputting the at least one object identifier and attributes of the at least one object identifier.

Abstract: Disclosed herein are apparatuses and methods for detecting a slip, trip or fall event in an environment and sending an alert of the event. An implementation may comprise detecting, in image frames captured by a camera, when an object enters a region of interest, tracking movements of the object, and determining that the object is in a fall-zone. The fall-zone may be specified by a set of line segments, each line segment being defined by points that lie in the region of interest. The implementation may further comprise, when the object is in the fall-zone, recording the position of the object, detecting a slip, trip or fall event when the object transitions from a first position above a predetermined height threshold to a second position below the predetermined height threshold, a pose of the object indicates that the object is lying down, or the object transitions from a vertical to horizontal pose, and sending the alert.

Abstract: A system may be configured to perform enhanced detection of occluded objects in a multiple object detection system. In some aspects, the system may detect a plurality of current objects in a current video frame, generate initial object identifier mapping information assigning a candidate object identifier to a first object of the plurality of current objects, and detect a multi-object tracking (MOT) inaccuracy condition. Further, the system may generate a current fingerprint corresponding to the first object of the plurality of current objects, identify an assignment error based on comparing the current fingerprint to historic fingerprints associated with a plurality of historic objects detected in a previous frame, generate updated object identifier mapping information based on the assignment error, and displaying the current video frame based on the updated object identifier mapping information.

Abstract: Disclosed herein are apparatuses and methods for auto-configuring a region of interest (ROI) associated with a camera. In one implementation, A method includes tracking, over a plurality of image frames spanning a first period of time, a plurality of objects moving in an environment, and determining a plurality of respective trajectories of the plurality of objects. The method includes identifying a common trajectory shared by more than one object of the plurality of objects based on comparing each of the plurality of respective trajectories to one another and identifying an ROI that encompasses the common trajectory. The method further includes tracking, over a plurality of new image frames spanning a second period of time, another plurality of objects moving in the environment, and generating an alert that the common trajectory has changed based on new trajectories of the another plurality of objects.

Abstract: Disclosed herein are apparatuses and methods for auto-configuring a region of interest (ROI) associated with a camera. In one implementation, a method comprises receiving image frames from a camera installed in the environment, wherein the ROI is located within the view of the image frames. The method includes tracking a plurality of persons in the image frames and determining a respective trajectory of movement for each person of the plurality of persons. The method further includes comparing each of the respective trajectories to one another and identifying, based on the comparing, a common trajectory shared by more than one person of the plurality of persons, wherein the common trajectory is not fully encompassed in the ROI. The method additionally includes updating the ROI to encompass the common trajectory, and includes configuring the updated ROI to be associated with new image frames from the camera.