Abstract: Methods, apparatus, systems, and computer-readable media are provided for using selectable elements to invoke an automated assistant at a computing device. While operating the computing device, a user may not be aware that the automated assistant can be invoked according to certain invocation phrases. In order to inform the user of the functionality of the automated assistant, the user can be presented with selectable elements that can initialize the automated assistant when selected. Furthermore, a selectable element can provide an invocation phrase in textual form so that the user is aware of their ability to invoke the automated assistant by speaking the invocation phrase. The selectable element can be presented at different devices associated with the user, and the automated assistant can be initialized at a device that is separate from the device where the selectable element is presented.

Abstract: The present disclosure is directed to generating animated images for presentation by a dynamic keyboard interface. In particular, the methods and systems of the present disclosure can: receive data describing advertisement content, and data describing a first context in which to present the advertisement content; generate data describing a first animated image including at least a portion of the advertisement content; determine a second context in which to present the advertisement content; generate data describing a second animated image including at least a portion of the advertisement content; and communicate, to one or more user devices on which one or more applications are executed, data indicating a plurality of different animated images for presentation by a dynamic keyboard interface in association with the one or more applications, the plurality of different animated images comprising the first animated image and the second animated image.

Abstract: A method can include receiving a starting scene for display and an ending scene for display, the starting scene including at least a first graphical element in a first location and a second graphical element in a second location, the ending scene including at least the first graphical element in a third location and the second graphical element in a fourth location; generating multiple individual candidate animations based on the starting scene and the ending scene, each of the multiple candidate animations including display of the first graphical element transitioning from the first location to the second location and display of the second graphical element transitioning from the third location to the fourth location; for each of the multiple individual candidate animations, determining a score; selecting one of the individual candidate animations based on the determined scores for the individual candidate animations; and presenting the selected individual candidate animation.

Abstract: An engagement time computer system for serving online content based on user engagement duration includes an output device, at least one memory, and one or more processors. The processors are configured to determine a historical engagement time of an online user and to identify a plurality of online content items. Each online content item of the plurality of online content items has a target engagement time. The processors are further configured to receive a request for a content item. The request is associated with the online user. The processors are also configured to select a first online content item from the plurality of online content items based at least in part on the historical engagement time of the online user and the target engagement time of the first online content item and to transmit the first online content item for display to the online user.

Abstract: A method can perform a process with a method including capturing an image, determining an environment that a user is operating a computing device, detecting a hand gesture based on an object in the image, determining, using a machine learned model, an intent of a user based on the hand gesture and the environment, and executing a task based at least on the determined intent.

Abstract: An audience measurement service may monitor broadcasts in a region, ingesting closed captioning data, metadata, displayed text or logos, performing facial recognition on people, and/or generating audio/video fingerprints of broadcast content to create a database of content and associated keywords, along with times at which the content was broadcast. The audience measurement service may receive user search queries from a search provider and detect “spiking” queries. The audience measurement service may determine whether the spiking queries are related, contextually and temporally, to an item of content identified in the database, and infer that users generating the queries likely viewed the item of content. In some implementations, the audience measurement service may also infer that the users likely viewed other items of content (e.g. advertisements) displayed during the prior or subsequent few minutes on the same broadcast channel as the query-triggering item.

Abstract: A system may provide, via a user interface, information identifying two or more of trending search-related information, trending video-related information, trending image-related information, or trending news-related information. The two or more of the trending search-related information, the trending video-related information, the trending image-related information, or the trending news-related information are visually presented as graphical items. Each graphical item, of the graphical items, is visually identified as corresponding to the trending search-related information, the trending video-related information, the trending image-related information, or the trending news-related information. The user interface presents multiple graphical items simultaneously. The system may further receive information identifying a topic, and provide, based on receiving the information identifying the topic, a subsequent group of graphical items, on the user interface.

Abstract: Techniques are described for training and/or utilizing an end-to-end speaker diarization model. In various implementations, the model is a recurrent neural network (RNN) model, such as an RNN model that includes at least one memory layer, such as a long short-term memory (LSTM) layer. Audio features of audio data can be applied as input to an end-to-end speaker diarization model trained according to implementations disclosed herein, and the model utilized to process the audio features to generate, as direct output over the model, speaker diarization results. Further, the end-to-end speaker diarization model can be a sequence-to-sequence model, where the sequence can have variable length. Accordingly, the model can be utilized to generate speaker diarization results for any of various length audio segments.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer-readable storage medium, and including a method for publishing content. The method comprises identifying a content item for publication. The method further comprises identifying one or more reviews associated with content included in the content item. The method further comprises evaluating a review including determining a rating for the content where determining the rating includes normalizing the rating to a first scale. The method further comprises extracting one or more relevant phrases from the review. The method further comprises publishing the content item along with a normalized rating and the one or more phrases.

Abstract: Systems and methods provides for dialog management in real-time rather than turn taking. An example method included generating first candidate responses to triggering event. The triggering event may be receipt of a live stream chunk for the dialog or receipt of a backend response to a previous backend request for a dialog shema. The method also includes updating a list of candidate responses that are accepted or pending with at least on of the first candidate responses, and determining, for the triggering event, whether the list of candidate responses includes a candidate response that has a confidence score that meets a triggering threshold. The method also includes waiting for a next triggering event without providing a candidate response when the list does not include a candidate response that has a confidence score that meets the triggering threshold.

Abstract: A system and method of displaying, on a touchscreen display of a mobile device, a virtual game controller having at least one virtual button configured to control multiple actions of an avatar representing a player within a video game executed by a remote gaming server. The player's interaction with the at least one virtual button is received through the touchscreen display and used to generate a control signal based on the received player interaction, the control signal being interpretable by the video game as an instruction to perform multiple actions in the video game. The control signal is transmitted from the mobile device for receipt by the remote gaming server executing the video game.

Abstract: The present disclosure provides an on-device machine learning platform that enables sharing of machine-learned models between applications on a computing device. For example, a first application which has a machine-learned model for a specific task can expose the model to other applications through a system level application programming interface (API) for the other applications to use. Communications using the API can be handled by the on-device machine learning platform. In some implementations, some exchange of resources (e.g., computing resources) can be provided so that the first application is compensated for sharing the machine-learned model (e.g., on a per model invocation basis).

Abstract: A wearable heads-up display (WHUD) obtains attribute data corresponding to an attribute of an item of interest and obtains environmental data of an environment surrounding the WHUD via one or more sensors of the WHUD. The WHUD compares the attribute data with the environmental data to detect the item of interest. In response to the detection, the WHUD obtains location data indicative of a location of the item of interest, stores the location data in association with a context of detection of the item of interest. In response to a trigger, such as a query by a user regarding the item of interest, the WHUD provides a location indication based on the location data, the location indication including, for example, a display of a description of the location of the item of interest, a display of the item of interest at the location, and the like.

Abstract: Automatically communicating information related to a user computing device from a central computing system to a transaction computing system prior to initiation of a transaction between the user computing device and the transaction computing system comprises a digital wallet application and a central computing system. A digital wallet application receives a beacon signal from a transaction computing system indicating the specific identification of the terminal with which the user intends to conduct a transaction. The user computing device communicates the terminal identification to the central computing system. The central computing system communicates customer data, such as loyalty information and offers, to the transaction computing system associated with the terminal. The transaction computing system communicates the customer data to the terminal. When the user taps the user computing device to initiate the transaction, the terminal provides any benefits associated with the customer data.

Abstract: Remote automated assistant component(s) generate client device notification(s) based on a received IoT state change notification that indicates a change in at least one state associated with at least one IoT device. The generated client device notification(s) can each indicate the change in state associated with the at least one IoT device, and can optionally indicate the at least one IoT device. Further, the remote automated assistant component(s) can identify candidate assistant client devices that are associated with the at least one IoT device, and determine whether each of the one or more of the candidate assistant client device(s) should render a corresponding client device notification.