Abstract: Systems and methods are disclosed herein for automatically identifying a query object within a visual medium. The technique generally involves receiving as input to a neural network a query object and a visual medium including the query object. The technique also involves generating, by the neural network, representations of the query object and the visual medium defining features of the query object and the visual medium. The technique also involves generating, by the neural network, a heat map using the representations. The heat map identifies a location of pixels corresponding to the query object within the visual medium and is usable to generate an updated visual medium highlighting the query object.

Abstract: Techniques for influencing power consumption of a client while streaming multimedia content from a server over a network are described. The power consumption is influenced by controlling how many requests are used by the client to receive media segments of the multimedia content from the server. For example, the number of requests is reduced by increasing the durations of the available media segments. Thus, rather than requesting a certain number of media segments each of a particular duration, a lower number of media segments is requested where the requested media segments have longer durations. Although a longer duration generally results in a larger power saving, the length of a requested media segment is balanced against other parameters associated with current conditions of the client, server, and/or network to allow for, for example, adaptive streaming. As such an optimum duration is determined based on different parameters associated with the current conditions.

Abstract: Various embodiments include automatically selecting and modifying objects within an image that are similar to an object previously selected by a user. A user first selects an object in a displayed image. Object feature descriptors are generated describing a shape formed by edges of the object, and image feature descriptors are generated describing shapes formed by edges automatically detected in the image. The object feature descriptors are compared with the image feature descriptors to automatically identify other objects in the image having edges similar to the selected object. The pixels of the user selected object and the automatically identified objects are modified according to a user specified function.

Abstract: User interface creation from screenshots is described. Initially, a user captures a screenshot of an existing graphical user interface (GUI). In one or more implementations, the screenshot is processed to generate different types of templates that are modifiable by users to create new GUIs. These different types of templates can include a snapping template, a wireframe template, and a stylized template. The described templates may aid GUI development in different ways depending on the type selected. To generate a template, the screenshot serving as the basis for the template is segmented into groups of pixels corresponding to components of the existing GUI. A type of component is identified for each group of pixels and locations in the screenshot are determined. Based on the identified types of GUI components and determined locations, the user-modifiable template for creating a new GUI is generated.

Abstract: Techniques are provided for k-best option identification of options subject to a supplied tolerance. One technique includes: sampling the options for a first period on a plurality of computers; computing an average and a sample count for each option based on the sampling; splitting the options into a highest group and a lowest group based on the computed averages; selecting a weakest one of the highest group (option A) and a strongest one of the lowest group (option B); and deciding whether or not to terminate based on the supplied tolerance and the selecting of options A and B. In some cases, the technique further includes outputting the highest group and terminating in response to a termination decision; otherwise continue with sampling options A and B for a next period; and updating the computed average and the sample count for options A and B based on corresponding next period sampling.

Abstract: Influence maximization determination within a social network system is described. In one example, a subset is selected from a plurality of user accounts of a social network system. Exposure of digital marketing content is then caused to the subset of user accounts. A determination is made as to a probability of each user account of the plurality of user accounts as being influenced by the exposure of the digital marketing content to the subset of user accounts. The determined probability is then output, such as to control output of digital marketing content.

Abstract: User data overlap determination in a digital medium environment is described. Initially, a user selects segments of user data for which a determination of overlap is to be made. For example, the user selects a segment representing users that are working professionals and a segment representing users that are mothers, such that working-mother users may correspond to the overlap. Regardless of the particular segments selected, an indication of those segments is received. One of multiple different overlap determining techniques—which can include a combined MinHash and HyperLogLog (HLL) technique and an Inclusion-Exclusion technique—may be selected for computing the overlap based on a number of segments indicated and numbers of users represented by the segments. The selected overlap determining technique is then used to compute the user data overlap between the indicated segments. Digital content including values indicative of the determined overlap is generated for presentation to a user.

Abstract: Slider manipulation with precision alteration is described. In example implementations, a computing device increases a precision level of a slider to make selecting a specific value easier and more efficient. Initially, a user interface is displayed that includes a slider having a first precision level with a first granularity of values. The computing device detects a first user interaction that moves a marker indicative of a value along the slider. Based on a detection of a second user interaction, the computing device alters the slider to have a second precision level with a second granularity of values. Examples of second user interactions include a movement to some distance away from the slider or a change in speed of the movement. By making the second granularity lower than the first granularity, the slider having the second precision level is displayed with wider separation spaces between adjacent values.

Abstract: Systems and methods are disclosed for analyzing a plurality of clickstreams associated with a resource to identify popular navigational patterns traversed by users of the resource. The analysis provides a navigational framework for performing continued analysis on segmented portions of the identified navigational patterns. To facilitate the analysis, clickstreams associated with the resource are analyzed to identify sets of clickstreams that have a common group of assets with which users of the resource interacted. Navigational patterns, which include commonly traversed series of assets interacted with by the users, are determined for the identified sets. The navigational pattern is then provided to identify popular navigational patterns traversed by users of the resource.

Abstract: An input device, such as a multifunction straight edge or a keyboard, has a recognizable contact shape when placed on a touchscreen display surface of a computing device. The contact shape of the input device can be a defined pattern of contact points, and a location and orientation of the input device on the touchscreen display surface is determinable from the defined pattern of the contact points. The input device includes an interaction module that interfaces with a companion module of the computing device. The companion module can initiate a display of an object responsive to the input device being recognized on the touchscreen display surface. The interaction module can receive a user input to the input device, and communicate the user input to the companion module of the computing device to modify the display of the object on the touchscreen display surface.

Abstract: Live manifest update techniques are described. In one or more implementations, data streams of media content are provided to computing devices by a service provider. During playback of the data streams, a manifest associated with the data streams is updated by the service provider effective to cause the computing devices to transition to playing back the data streams using the updated manifest without restarting playback of the data streams. In one or more implementations, a data stream of media content is played back at a computing device. During playback of the data steam, a manifest monitoring module at the computing device monitors for updates to a manifest associated with the data stream. Responsive to detecting an update to the manifest, the computing device seamlessly transitions to playing back the data stream using the updated manifest without restarting playback of the data stream.

Abstract: Systems and methods for are disclosed for providing integrated display of data metrics from different data sources. In some embodiments, an icon representing a de-limited data file can be positioned on a graphical interface that displays first metrics data for a first time period. An integration module can respond to the positioning of the icon by determining that the de-limited data file includes second metrics data for a second time period at least partially overlapping with the first time period. The integration module can update the graphical interface to display at least some of the second metrics data from the delimited data file with the first metrics data. The first metrics data and the second metrics data can be displayed for at least an overlapping portion between the first and second time periods.

Abstract: Techniques to facilitate image searches using context of an image frame in a document are described. In one or more implementations, an image frame included within a document is detected, such as when the frame is inserted or selected. Content in the document in the vicinity of the image frame is analyzed to derive query terms corresponding to the content. Tag elements created for the query terms are exposed to enable selection of individual query terms as search parameters. An image search is conducted using selected query terms and images returned as results of the search are presented a candidates for insertion into the frame. Previews of images as being inserted in the image frame can be exposed show how the document would look with different images placed in the frame. Selection of a particular image from the arrangement causes insertion of the particular image into the image frame.

Abstract: One exemplary embodiment involves a computer-implemented method comprising receiving, in a computing device, a first request to attach a file in a cell of a collaborative spreadsheet, wherein the first request comprises the file and a cell identifier, and wherein the collaborative spreadsheet comprises a spreadsheet configured to be accessed and modified simultaneously from a plurality of client devices. The method additionally comprises attaching, via a processor, the file to the collaborative spreadsheet at the cell corresponding to the cell identifier; and storing the file on a cloud storage device, wherein the file is stored to be accessible in conjunction with the collaborative spreadsheet.

Abstract: Systems and methods disclosed herein offer assistance to customers shopping at a real-world shopping venue carrying mobile devices connected to a communication network. This involves tracking a physical action of a customer present at the real-world shopping venue and determining to offer assistance to the customer based on the tracked physical action. Based on determining to offer assistance to the customer, a notification is sent to a mobile device of the customer. The notification identifies that live support is available to assist the customer while the customer is present at the real-world shopping venue.

Abstract: Techniques are disclosed for automatically detecting anomalies in the content of a scheduled social media post, alerting a user to the presence of such anomalies before the content is posted and recommending a course of action when an anomaly is detected. A set of keywords is extracted from a scheduled post using an ontological classification technique. At predetermined time intervals, the keywords are compared with information obtained from one or more data sources to determine if an anomaly is present. If an anomaly is detected, the scheduled post is classified into one of at least three categories: supporting the post, neutral, or opposing the post. Once the anomaly is detected and the scheduled post is classified, the author of the post is alerted to the anomaly along with the categorization. Subsequently, the author may reschedule the post to an earlier or later time, delete the post, or change the post.

Abstract: Request-driven file pulling from unsynchronized directories is described. Rather than maintaining entire files, a service provider merely maintains file tags which describe the files in unsynchronized directories of devices associated with a user and communicates the file tags to the devices. The file tags enable indications of the files to be presented so that selected files can be pulled from the unsynchronized directories. Based on user selections made at a first of the devices to pull files from a second of the devices, requests are communicated from the first device to the service provider for the selected files. Responsive to a request, the service provider communicates an indication to the second device to copy the selected file from the second device's unsynchronized directory to a synchronized directory. Responsive to this, the service provider obtains the file from the synchronized directory and delivers the file to the first device.

Abstract: Digital content translation techniques and system are described. In one example, source digital content is linked via metadata to different derived format versions that are generated from the source digital content. The metadata, for instance, may be used to locate source digital content that generated a particular derived format version that is in use by a service provider system. The source digital content, once identified and located, may then be used to improve efficiency and accuracy in translation of text or images included as part of the source digital content. The updated source digital content is then used to generate a derived format version that includes the translated text or other portion, e.g., an image.

Abstract: Digital asset association techniques with search query data are described. In one example, A first digital asset is displayed at a location within digital content in a user interface. The location is specified using placeholder data of the digital content. An input is received via selection of an option as part of the user interface to initiate a search. Search query data associated with the placeholder data is obtained in response to the input. A search is initiated for a second digital asset using the search query data. The second digital asset is displayed at the location within the digital content in the user interface as specified by the placeholder data.

Abstract: Probabilistic determination of selected image portions is described. In one or more implementations, a selection input is received for selecting a portion of an image. For pixels of the image that correspond to the selection input, probabilities are determined that the pixels are intended to be included as part of a selected portion of the image. In particular, the probability that a given pixel is intended to be included as part of the selected portion of the image is determined as a function of position relative to center pixels of the selection input as well as a difference in one or more visual characteristics with the center pixels. The determined probabilities can then be used to segment the selected portion of the image from a remainder of the image.