Abstract: Methods, and systems including computer programs encoded on a computer storage medium, for training a detection model for surveillance devices using semi-supervised learning. In one aspect, the methods include receiving imaging data collected by a camera of a scene within a field of view of the camera. Annotated training data is generated from the imaging data and one or more detection models are trained using the annotated training data. Based on a set of performance parameters, an optimized detection model is selected of the one or more detection models, and the optimized detection model is provided to the camera.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an event detector. The methods, systems, and apparatus include actions of obtaining frames of a video, determining whether an object of interest is detected within the frames, determining whether motion is detected within the frames, determining whether the frames correspond to motion by an object of interest, generating a training set that includes labeled inter-frame differences based on whether the frames correspond to motion by an object of interest, and training an event detector using the training set.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for multi-query object matching based on inverse model frequency. The methods, systems, and apparatus include actions of obtaining images of a sample object, obtaining models of known objects, determining an image score for each pair of the images and the models, determining an inverse model frequency for each image based on the image scores, determining a model score for each model based on the inverse model frequencies and the image scores, and selecting a particular known object of the known objects as the sample object based on the model scores.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for monitoring system with a video analytic enhanced pet door. The methods, systems, and apparatus include actions of obtaining video from a camera, determining from the video that an object is approaching the camera, in response to determining from the video that an object is approaching the camera, determining whether the object is an animal, in response to determining that the object is an animal, determining from the video whether the animal is authorized to enter a home, and in response to determining from the video that the animal is authorized to enter the home, opening a pet door.

Abstract: An electronic peep hole device includes a first portion that is configured to be mounted to an exterior side of a front door of a monitored property, the first portion includes a user interface and a camera that is configured to capture image data, a second portion that is configured to be mounted to an interior side of the front door of the monitored property, the second portion includes a display unit that is configured to display the image data captured by the camera, and a connecting portion that is configured to connect the first portion to the second portion through a hole in the surface of the front door, the connecting portion being configured to house an electrical connection between the first portion and the second portion.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for classifying actions of an object. The methods, systems, and apparatus include actions of: obtaining frames of video including an object of interest; determining a type of action of the object in each of the frames of video; determining a group of frames from the frames of video based on the type of action; determining an aggregated background subtraction (ABS) image based on adjacent frames of the group of frames; generating a training set that includes labeled ABS images including the ABS image; and training an action classifier using the training set.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for obtaining a foreground occupancy map for a camera view. The methods, systems, and apparatus include actions of determining an area of an image in which there is a false detection of an object, determining a likely contribution of the area to the false detection based on the foreground occupancy map, generating a modified object detector based on the likely contribution of the area, and detecting an object using the modified object detector.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for machine learning for home understanding and notification. In one aspect, a method includes obtaining reference videos from a camera within a home, determining from the reference videos that a particular person routinely leaves the home with a particular object at a particular time of day, determining from a sample video from the camera within the home that the particular person appears to be leaving the home without the particular object at the particular time of day, and in response, providing a notification regarding the particular object.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for synchronizing recorded videos. The methods, systems, and apparatus include actions of receiving a first video and a second video, determining first time stamps from pixels in the first video, determining second time stamps from pixels in the second video, determining an earliest frame in each of the videos that time stamps match, pairing the earliest frames, determining that a subsequent frame in each of the videos have corresponding time stamps that have a difference that satisfies a criteria, and pairing the subsequent frame in the first video with a frame after the subsequent frame in the second video.

Abstract: A method and apparatus for determining a building/urban profile image location includes extracting a building/urban profile from an obtained image and comparing the image to a database of building/urban profiles. The database may be created by obtaining point cloud data sets scanned building/urban profiles, converting the obtained point cloud data sets to a corresponding 3D surface model, and creating the database of stored building/urban profiles viewed from multiple locations within each of the 3D surface models. By comparing the extracted building/urban profile image with the stored building/urban profiles contained within the database and finding a match the location from which the extracted building/urban profile image was taken can be determined.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for detecting humans in images using occupancy grid maps. The methods, systems, and apparatus include actions of obtaining an image of a scene without people, generating a reference occupancy grid from the image, generating pairs of training images with humans rendered and corresponding training occupancy grids based on the occupancy grid and the image, training a scene-specific human detector with the pairs of training images with humans rendered and corresponding training occupancy grids, generating a sample occupancy grid from a sample image using the scene-specific human detector, and augmenting the sample image using the sample occupancy grid.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for monitoring wildlife using a self-powered camera. The methods, systems, and apparatus include actions of detecting an object near a camera, in response to detecting the object near the camera, classifying the object as a particular type of animal based on an appearance of the object in an image captured by the camera, determining whether to provide a notification regarding the image based on the particular type of animal that the object is classified as, and in response to determining to provide a notification regarding the image based on the particular type of animal that the object is classified as, providing the notification regarding the image.

Abstract: A computer-implemented method includes obtaining data from one or more non-visual sensors and a camera from a first monitoring system. The data includes non-visual data from the non-visual sensors and visual data obtained from the camera. The non-visual data from the non-visual sensors are paired with corresponding visual data from the camera. Data points of the non-visual data are synchronized with frames of the visual data based on a likelihood of an event indicated in the non-visual data. The synchronized data points of the non-visual data with the frames of the visual data are provided as labeled input to a neural network to train the neural network to detect the event. The trained neural network is provided to one or more cameras corresponding to one or more additional monitoring systems to detect the event in the visual data obtained by the one or more cameras.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a storage device, for monitoring a vehicle is disclosed. In one aspect, a vehicle monitoring system is disclosed that includes a processor and a storage device storing instructions that, when executed by the processor, causes the processor to perform operations. The operations may include obtaining data generated by a vehicle monitoring component, determining whether the parked vehicle is permitted to navigate away from a location where the vehicle is parked, and in response to determining that the parked vehicle is not permitted to navigate away from the location, providing alert data to a processing unit installed within the parked vehicle, wherein processing of the generated alert data by the processing unit triggers a notification that is configured to alert the vehicle operator that the parked vehicle is not permitted to navigate away from the location.

Abstract: A monitoring system that is configured to monitor a includes a light device that is located at a pool of the property and that includes a camera that is configured to capture video data, and a monitoring control unit that is configured to receive, from the light device, the video data, determine that an object in the pool of the property is likely a person, based on determining that an object in the pool of the property is likely a person, determine a confidence score that reflects a likelihood that the person is swimming, determine that the confidence score does not satisfy a confidence score criteria, and perform a monitoring system action based on determining that the confidence score does not satisfy the confidence score criteria.

Abstract: A vent hood device that is configured to monitor a stove, the vent hood device including, a camera, an infrared sensor that is configured to detect heat from a surface of the stove, and a processor that is configured to receive, from the infrared sensor, temperature data that indicates heat detected from the surface of the stove, receive, from the camera, video data captured by the camera, determine, based on the temperature data and the video data, that an adult is not within a field of view of the camera and the heat detected from the surface of the stove satisfies a heat criteria, and in response to determining that an adult is not within the field of view of the camera and the heat detected from the surface of the stove satisfies the heat criteria, provide an alert that indicates that the stove is unattended.

Abstract: A computer-implemented method includes obtaining an image of an area of a property from an augmented reality device, identifying the area of the property based on the image obtained from the augmented reality device, determining that the area of the property corresponds to an event at the property or a configuration of a monitoring system of the property, and providing, in response to determining that the area of the property corresponds to the event or the configuration, information that represents the event or the configuration and that is configured to be displayed on the augmented reality device.

Abstract: Methods and systems including computer programs encoded on a computer storage medium, for obtaining imaging data of a parking lot that includes a set of parking spots, detecting a vehicle enter the parking lot, generating a vehicle recognition model for the vehicle, determining that the vehicle is parked in a parking spot, detecting a customer exit the vehicle, generating a customer recognition model for the customer, determining that the customer has entered a business of one or more businesses affiliated with the parking lot, and providing information related to parking space usage for the parking spot and the business to a user device.

Abstract: Methods and apparatus are disclosed for enhancing urban surface model with image data obtained from a satellite image. Three dimensional models of an urban cityscape obtained from digital surface models may comprise surface location information but lack image information associated with the cityscape, such as the color and texture of building facades. The location of the satellite at the time of recording the satellite image interest may be obtained from metadata associated with the satellite image. A 3D model of a cityscape corresponding to the satellite image may be subjected to a transformation operation to determine portions of the 3D model that are viewable from a location corresponding to the location of the satellite when taking the picture. Visible facades buildings of the 3D model may be identified and mapped to portions of the satellite image which may then be used in rendering 2D images from the 3D model.

Abstract: A method and apparatus for determining a building/urban profile image location includes extracting a building/urban profile from an obtained image and comparing the image to a database of building/urban profiles. The database may be created by obtaining point cloud data sets scanned building/urban profiles, converting the obtained point cloud data sets to a corresponding 3D surface model, and creating the database of stored building/urban profiles viewed from multiple locations within each of the 3D surface models. By comparing the extracted building/urban profile image with the stored building/urban profiles contained within the database and finding a match the location from which the extracted building/urban profile image was taken can be determined.