Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for converting unstructured documents to structured key-value pairs. In one aspect, a method includes: providing an image of a document to a detection model, wherein: the detection model is configured to process the image to generate an output that defines one or more bounding boxes generated for the image; and each bounding box generated for the image is predicted to enclose a key-value pair including key textual data and value textual data, wherein the key textual data defines a label that characterizes the value textual data; and for each of the one or more bounding boxes generated for the image: identifying textual data enclosed by the bounding box using an optical character recognition technique; and determining whether the textual data enclosed by the bounding box defines a key-value pair.

Abstract: Methods, systems and apparatus, including computer program products are described for providing temporally relevant targeted content items. A set of user data is used to determine current user interests as indicated by one or more recent events. Content items relevant to those events are identified and ranked based on a time that the corresponding event occurred. One or more of the top ranked content items can be delivered to a user.

Abstract: Methods, systems, and media for presenting caption information based on volume setting adjustments are provided.

Abstract: A system and method are provided that permits different devices with different applications to respond to the occurrence of a condition in a similar manner. By way of example, trigger data may be synchronized across multiple devices, wherein trigger data identifies information to be provided to a program in response to the occurrence of a condition. If a program is not installed or running on a particular device when the trigger occurs, the device automatically installs and executes the application.

Abstract: A computing unit is disclosed, comprising a first memory bank for storing input activations and a second memory bank for storing parameters used in performing computations. The computing unit includes at least one cell comprising at least one multiply accumulate (“MAC”) operator that receives parameters from the second memory bank and performs computations. The computing unit further includes a first traversal unit that provides a control signal to the first memory bank to cause an input activation to be provided to a data bus accessible by the MAC operator. The computing unit performs one or more computations associated with at least one element of a data array, the one or more computations being performed by the MAC operator and comprising, in part, a multiply operation of the input activation received from the data bus and a parameter received from the second memory bank.

Abstract: A method includes receiving an input video stream and scaling the input video stream into two or more spatial layers. For each spatial layer, the method also includes generating a temporal layer prediction pattern by: obtaining a temporal base layer for a corresponding spatial layer; identifying, based on the temporal base layer, a plurality of temporal layers and a plurality of temporal time slots during a temporal period; and aligning the temporal base layer for the corresponding spatial layer with one of the temporal time slots during the temporal period. Each temporal time slot is associated with one of the temporal base layer or one of the plurality of temporal layers for the corresponding spatial layer. The temporal base layer for each corresponding spatial layer is aligned with a different temporal time slot than each other temporal base layer for each other corresponding spatial layer.

Abstract: A system includes a chassis. The system also includes a panel extending in a plane, the panel having a first edge and a second edge opposite the first edge. The system further includes an actuator coupled to the panel at a first location, the actuator being adapted to cause vibration of the panel to generate sound waves. The system further includes a coupling assembly mechanically coupling the chassis to the panel at the second edge of the panel. The panel is fixedly connected to the chassis at the first edge to limit displacement of the panel from the chassis at the first edge during vibration of the panel by the actuator, and the coupling assembly allows displacement of the panel from the chassis at the second edge during the vibration of the panel by the actuator.

Abstract: Systems and methods according to aspects of the invention verify requests for information over a network. Computing devices communicating over a network transmit numerous signals with various types of information including spatial information, temporal information, device-specific information, and transmission specific information. This information may be used to generate statistical records of valid request for information. Various combinations of these signals may be used to determine whether a request for information is a valid request based on the statistical records associated with the signals. Where a request for information has been verified, the system and method may provide the information. Where the request for information has not been verified, the system and method may deny the request for information. Thus the system and method may be used to prevent and block scraping and spamming attacks.

Abstract: A computer system stores information for a plurality of layouts of one or more user interface elements locally. Each layout corresponds to at least one type of display device, and includes one or more objects associated with the one or more user interface elements. Upon receiving a request for rendering a first user interface element on a display device, the computer system identifies a type of the display device and determines a first one of the at least one type of display device accordingly. The computer system then extracts information for a first layout corresponding to the first one of the at least one type of display device (e.g., information for a subset of objects that is stored with the first layout and corresponds to the first user interface element), and renders the subset of objects corresponding to the first user interface element on the display device accordingly.

Abstract: This document describes techniques and devices for a radar recognition-aided search. Through use of a radar-based recognition system, gestures made by, and physiological information about, persons can be determined. In the case of physiological information, the techniques can use this information to refine a search. For example, if a person requests a search for a coffee shop, the techniques may refine the search to coffee shops in the direction that the person is walking. In the case of a gesture, the techniques may refine or base a search solely on the gesture. Thus, a search for information about a store, car, or tree can be made responsive to a gesture pointing at the store, car, or tree with or without explicit entry of a search query.

Abstract: An interactive multi-view module identifies a plurality of media items associated with one or more real-world events. Each of the plurality of media items is created by capturing the real-world events from a particular geographic location. The interactive multi-view module determines a geographic position associated with each of the media items and presents the media items in an interactive multi-view player interface based at least on the geographic positions. The interactive multi-view player interface allows a user to switch between of media items, and indicates at least one of a video density indicating a number of media items available at a given point in time or an event highlight indicating a popularity of the media items at a given point in time. The popularity of the respective media items is determined using one or more factors comprising a number of views of the media items at a given point in time.

Abstract: Systems and techniques are provided for determination of user presence and absence using WiFi connections. Reports may be received from WiFi access points in an environment. The reports may include an identifier of a WiFi device, an indication of a connection to or disconnection from a WiFi access point, a time of the connection or disconnection, and an identifier of the WiFi access point. A connection sequence for the WiFi device may be generated from the reports. Whether the WiFi device is present in or absent from the environment as of a specified time may be determined based on the connection sequence. An indication of presence for a user associated with the WiFi device may generated if the WiFi device is present in the environment. An indication of absence for the user associated with the WiFi device may be generated if the WiFi device is absent from the environment.

Abstract: The present disclosure includes systems and techniques relating to ranking search results of a search query. In general, the subject matter described in this specification can be embodied in a computer-implemented method that includes determining a measure of relevance for a document result within a context of a search query for which the document result is returned, the determining being based on a first number in relation to a second number, the first number corresponding to longer views of the document result, and the second number corresponding to at least shorter views of the document result; and outputting the measure of relevance to a ranking engine for ranking of search results, including the document result, for a new search corresponding to the search query. The subject matter described in this specification can also be embodied in various corresponding computer program products, apparatus and systems.

Abstract: A method includes identifying a plurality of media content items for a playlist, each of the plurality of media content items having an introductory segment, one or more main segments, and an ending segment; determining a length of an ending segment of a first media content item and a length of an introductory segment of a second media content item; selecting an interstitial to be added to the playlist between one or more main segments of the first media content item and one or more main segments of the second media content item; and responsive to determining that a length of the interstitial is less than a combined length of the ending segment and the introductory segment, adjusting an overlap on the interstitial of the ending segment of the first content media item and the introductory segment of the second media content item.

Abstract: Methods, systems, and media for presenting related media content items are provided.

Abstract: Systems and methods are disclosed for identifying viewing characteristics of an audience of a content channel. A method includes receiving an identification of a channel of interest having a plurality of first videos and being associated with a first channel owner in a network-based media sharing platform that provides a plurality of watch sessions to a plurality of users. The plurality of watch sessions comprises a subset of watch sessions that each include a first watched video from the channel of interest and one or more of a plurality of second watched videos from a plurality of additional channels, each of the plurality of additional channels associated with one of a plurality of second channel owners.

Abstract: Methods for receiving a request to process, on a hardware circuit, a neural network comprising a first convolutional neural network layer having a stride greater than one, and in response, generating instructions that cause the hardware circuit to, during processing of an input tensor, generate a layer output tensor equivalent to an output of the first convolutional neural network layer by processing the input tensor using a second convolutional neural network layer having a stride equal to one but that is otherwise equivalent to the first convolutional neural network layer to generate a first tensor, zeroing out elements of the first tensor that would not have been generated if the second convolutional neural network layer had the stride of the first convolutional neural network layer to generate a second tensor, and performing max pooling on the second tensor to generate the layer output tensor.

Abstract: A method includes receiving, from a multiscopic image capture system, a plurality of images depicting a scene. The method includes determining, by application of a neural network based on the plurality of images, a disparity map of the scene. The neural network includes a plurality of layers, and the layers include a rectification layer. The method include determining a matching error of the disparity map based on differences between corresponding pixels of two or more images associated with the disparity map. The method includes back-propagating the matching error to the rectification layer of the neural network. Back-propagating the matching error includes updating one or more weights applied to the rectification layer.

Abstract: According to an aspect, a method includes installing an application on a computing device, determining whether to activate a compatibility mode for the application, in response to activating the compatibility mode, determining a restriction to a change to an application window size or shape of a user interface of the application, and rendering a user interface object defining a plurality of predefined sizing options for the user interface of the application.

Abstract: Methods, systems and apparatus for quantum error correction. A layered representation of error propagation through quantum error detection circuits is constructed. The layered representation includes multiple line circuit layers that each represent a probability of local detection events in a quantum computing system associated with potential error processes in an execution of a quantum algorithm. To construct the layered representation, potential detection events associated with each potential error process occurring at quantum gates in the quantum circuit are determined. Lines are associated with each potential error process, the lines each connecting a potential detection event associated with the potential error process to another potential detection event associated with the same potential error process or a boundary of the quantum circuit. Similar lines are merged and used to construct unique line circuit layers.