Various techniques are provided for generating reports of three dimensional (3D) images. The techniques include identifying a plurality of volume features in a 3D image using a first machine learning (ML) module trained with annotated 3D images, and identifying a plurality of semantic representations associated with the 3D image using a second ML module trained with the annotated 3D images and reports associated with the annotated 3D images. The techniques further include generating a report of the 3D image based on the volume features and the semantic representations using a third ML module trained with the reports and outputs generated by the first ML module and the second ML module using the annotated 3D images and the reports.

Methods, systems, and apparatus, including computer programs encoded on computer storage media relate to a method for providing video communications with dynamic virtual backgrounds within a communication platform. Based on the contextual information of a meeting, the system selects a virtual background to be used for the meeting. During, a video meeting, the system generates for display, on one or more client devices, a composite video depicting the imagery of the user overlaid on imagery of the selected virtual background.

Provided is a method, computer program product, and system for building an object checklist used to predict which objects should accompany a user during an occasion. A processor may monitor observation data related to the user from an observation device. The processor may analyze the observation data to identify an object associated with the user. The processor may collect contextual data related to the identified object and the user. The processor may compare the contextual data to a usage pattern threshold related to the identified object and one or more other objects. In response to the usage pattern threshold being met, the processor may output an alert to the user. The alert may indicate that an exception to a usage pattern related to the identified object has occurred.

Aspects of the subject disclosure may include, for example, a device having a processing system including a processor; and a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations including capturing images generated by a fiducial invoked on a user device; determining a context of the fiducial; detect an anomaly in the images based on the context; and responsive to detecting the anomaly, providing a notification of the anomaly. Other embodiments are disclosed.

Method, system and media for authenticating a subject as a user. Embodiments generate visual stories specific to the user and for which the subject must select the corresponding images from among a plurality of decoy images. Gaze tracking can be used to determine which images the user has selected without allowing an observer to learn which images have been selected. Images for the visual story can be retrieved from the user's social networking profile, and corresponding text storied generated to indicate which images should be selected. Multiple security levels are possible by varying the number of story images and decoy images.

One or more image parameters of an image may be analyzed using a hierarchical set of models. Executing individual models in the set of models may generate outputs from analysis of different image parameters of the image. Inputs of one or more of the models may be conditioned on a set of outputs derived from one or more preceding model in the hierarchy.

A method of detecting and responding to a visitor to a smart home environment via an electronic greeting system of the smart home environment, including determining that a visitor is approaching an entryway of the smart home environment; initiating a facial recognition operation while the visitor is approaching the entryway; initiating an observation window in response to the determination that a visitor is approaching the entryway; obtaining context information from one or more sensors of the smart home environment during the observation window; and at the end of the time window, initiating a response to the detected approach of the visitor based on the context information and/or an outcome of the facial recognition operation.

The present disclosure relates to a method and system of processing original handwriting input, the system and method being capable of recognize a plurality of strokes provided on the input recognition interface, the method including: determining a stroke box around each stroke; determining overlap between the stroke boxes; correlating overlapping stroke boxes to one or more characters; providing a character box around each of the one or more characters; determining overlap between character boxes; correlating overlapping character boxes to one or more words; providing a word box around each of the one or more words; provide a word margin around each of the one or more word boxes; determining overlap between each word box to determine a line; wherein each of the characters, words, or lines can be individually selected and rearranged, the system automatically adjusting spacing or placement of surrounding elements to allow for the rearrangement.

Various embodiments herein each include at least one of systems, methods, software, and data structures for context-aided machine vision. For example, one method embodiment includes identifying a customer in a shopping area and maintaining an item bin in a computing system of data identifying items the customer has picked up for purchase. This method further includes receiving an image of the customer holding an item and performing item identification processing on the image to identify the item the customer is holding. The item identification processing may be performed based in part on a stored shopping history of the customer indicating items the customer is more likely to purchase. The identified item is then added to the item bin of the customer.

Systems and methods for processing speech are described. In certain examples, image data is used to generate visual feature tensors and audio data is used to generate audio feature tensors. The visual feature tensors and the audio feature tensors are used by a linguistic model to determine linguistic features that are usable to parse an utterance of a user. The generation of the feature tensors may be jointly configured with the linguistic model. Systems may be provided in a client-server architecture.