Abstract: Methods and systems are directed to improving the convenience of drawing applications. Some examples include generating 3D drawing objects using a drawing application and selecting one based on a 2D design (in some cases a hand-drawn sketch) provided by a user. The user provided 2D design is separated into an outline perimeter and interior design, and corresponding vectors are then generated. These vectors are then used with analogous vectors generated for drawing objects. The selection of a drawing object to correspond to the 2D design is based on finding a drawing object having a minimum difference between its vectors and the vectors of the 2D design. The selected drawing object is then used to generate a drawing object configured to receive edits from the user. This reduces the inconvenience required to manually reproduce the 2D design in the drawing application.

Abstract: Font recognition and similarity determination techniques and systems are described. In a first example, localization techniques are described to train a model using machine learning (e.g., a convolutional neural network) using training images. The model is then used to localize text in a subsequently received image, and may do so automatically and without user intervention, e.g., without specifying any of the edges of a bounding box. In a second example, a deep neural network is directly learned as an embedding function of a model that is usable to determine font similarity. In a third example, techniques are described that leverage attributes described in metadata associated with fonts as part of font recognition and similarity determinations.

Abstract: Stroke operation prediction techniques and systems for three-dimensional digital content are described. In one example, stroke operation data is received that describes a stroke operation input via a user interface as part of the three-dimensional digital content. A cycle is generated that defines a closed path within the three-dimensional digital content based on the input stroke operation and at least one other stroke operation in the user interface. A surface is constructed based on the generated cycle. A predicted stroke operation is generated based at least in part on the constructed surface. The predicted stroke operation is then output in real time in the user interface as part of the three-dimensional digital content as the stroke operation data is received.

Abstract: Multi-video registration for video synthesis is described. In example implementations, at least one computing device synthesizes multiple videos to create merged images using an automated mechanism to register the multiple videos. The computing device obtains multiple videos with each video including a sequence of multiple frames. Using multiple camera poses determined in a three-dimensional scene reconstruction, respective frames of respective ones of the multiple videos are linked to produce linked frames. The computing device aligns the linked frames to produce aligned frames using point guidance that is based on the multiple spatial points identified in the 3D scene reconstruction. For example, pixels in each of the linked frames that correspond to a same spatial point of the three-dimensional scene reconstruction can be used to align the linked frames at a pixel level. Based on the aligned frames, the computing device creates at least one merged image to synthesize the multiple videos.

Abstract: Systems and methods provide for data management and governance to enforce proper usage of combined data formed from different pieces of data. Data governance metadata identifying authentication states for the individual pieces of data are combined to tag the combined data with new data governance metadata indicating the cumulative authentication states. Similarly, data governance metadata identifying standard contractual considerations for the individual pieces of data are combined to tag the combined data with new data governance metadata indicating the cumulative standard contractual considerations. An extent to which the combined data can be used to identify and/or contact a person is determined, and data governance metadata is tagged to the combined data identifying the extent to which the data can be used to identify and/or contact the person.

Abstract: Apparatus, systems, and methods may operate to determine a proximate save time difference between a target version of electronic content and a next (proximate in time and newer) version of the electronic content. An aged save time difference may be determined as the time difference between the current time and either one of the target version or the next version. The target version of the electronic content may be pruned from a plurality of revisions of the electronic content if the proximate save time difference is less than a selected fraction of the aged save time difference. The plurality of revisions, less the pruned revisions, may be displayed along a historical time line as part of a graphical user interface.

Abstract: The present disclosure relates dialog states, which computers use to internally represent what users have in mind in dialog. A dialog state tracker employs various rules that enhance the ability of computers to correctly identify the presence of slot-value pairs, which make up dialog states, in utterances or conversational input of dialog. Some rules provide for identifying synonyms of values of slot-values pairs in utterances. Other rules provide for identifying slot-value pairs based on coreferences between utterances and previous utterances of dialog sessions. Rules are also provided for carrying over slot-value pairs from dialog states of previous utterances to a dialog state of a current utterance. Yet other rules provide for removing slot-value pairs from candidate dialog states, which are later used as dialog states of utterances.

Abstract: Multi-video registration for video synthesis is described. In example implementations, at least one computing device synthesizes multiple videos to create merged images using an automated mechanism to register the multiple videos. The computing device obtains multiple videos with each video including a sequence of multiple frames. Using multiple camera poses determined in a three-dimensional scene reconstruction, respective frames of respective ones of the multiple videos are linked to produce linked frames. The computing device aligns the linked frames to produce aligned frames using point guidance that is based on the multiple spatial points identified in the 3D scene reconstruction. For example, pixels in each of the linked frames that correspond to a same spatial point of the three-dimensional scene reconstruction can be used to align the linked frames at a pixel level. Based on the aligned frames, the computing device creates at least one merged image to synthesize the multiple videos.

Abstract: Customizable reproduction of electronic meetings is described. In one or more embodiments, content, such as slides or audio or chat, is obtained from multiple participants of an electronic meeting, such as a web conference. Obtained content is channelized by associating a unique identifier with content based at least partially on which participant contributed the content. Content events may be associated with a time stamp that is derived from a time index of the electronic meeting. Content may be contributed by a presenter participant or a non-presenter participant. To customize reproduction of an electronic meeting, a host may be empowered to select content by designating specific participants for inclusion or exclusion in the reproduction. A user interface may graphically indicate types or time slots of content contributed by individual participants. A host may designate individual content events for inclusion or exclusion by selection or non-selection of displayed time slots.

Abstract: Digital experience content personalization and recommendation techniques within an AR or VR environment are described. In one example, a user profile is received that models how user interaction occurs with respect to virtual objects within a virtual or augmented reality environment. Digital experience content is obtained that defines a virtual or augmented reality environment. A virtual object is selected for inclusion as part of the digital experience content based at least in part on the user profile. Digital experience content is generated to support user interaction with the selected virtual object as part of the virtual or augmented reality environment.

Abstract: A digital medium environment is described to control provision of digital content within a physical environment to a mobile device associated with a user. User identification data and position data are received. The position data describes a physical location at which the mobile device is disposed within the physical environment. A user profile is selected based on the user identification data. The user profile describes user online interaction with digital content. Digital content is generated that is personalized based on the selected user profile and the position data. Output is then controlled of the generated digital content to the mobile device.

Abstract: Disclosed are techniques for more accurately estimating the pose of a camera used to capture a three-dimensional scene. Accuracy is enhanced by leveraging three-dimensional object priors extracted from a large-scale three-dimensional shape database. This allows existing feature matching techniques to be augmented by generic three-dimensional object priors, thereby providing robust information about object orientations across multiple images or frames. More specifically, the three-dimensional object priors provide a unit that is easier and more reliably tracked between images than a single feature point. By adding object pose estimates across images, drift is reduced and the resulting visual odometry techniques are more robust and accurate. This eliminates the need for three-dimensional object templates that are specifically generated for the imaged object, training data obtained for a specific environment, and other tedious preprocessing steps.

Abstract: Enhanced vectorization of raster images is described. An image vectorization module converts a raster image with bitmapped data to a vector image with vector elements based on mathematical formulas. In some embodiments, spatially-localized control of a vectorization operation is provided to a user. First, the user can adjust an intensity of a denoising operation differently at different areas of the raster image. Second, the user can adjust an automated segmentation by causing one segment to be split into two segments along a zone marked with an indicator tool, such as a brush. Third, the user can adjust an automated segmentation by causing two segments to be merged into a combined segment. The computation of the vector elements is based on the adjusted segmentation. In other embodiments, semantic information gleaned from the raster image is incorporated into the vector image to facilitate manipulation, such as joint selection of multiple vector elements.

Abstract: Contextual linking of digital representations of objects with documents is described. In one or more implementations, a portion of a document is linked to a portion of a digital representation of an object. The link contains information such as portion locations/coordinates, customized views, hyperlinks, text/notes associated with the link, audio/video associated with the link, and so on. The resulting linked document enables a user to freely navigate between the linked portions through user inputs to either portion and interact with the information associated with the link.

Abstract: Computerized methods and systems remove background information from digitally encoded images. A digitally encoded image is retrieved and converted to greyscale if it is encoded as a color image. The greyscale image is divided into a first set of subsections. The first set of subsections is processed to individually remove any background portions from each subsection of the first set of subsections. The greyscale image may also be divided into a second set of subsections and the second set of subsections is processed to individually remove any background portions from each subsection of the second set of subsections containing contours. The first set of subsections and second set of subsections may be merged to create a new version of the image, and the new version of the image is stored to digital storage.

Abstract: Embodiments described herein are directed to methods and systems for facilitating control of smoothness of transitions between images. In embodiments, a difference of color values of pixels between a foreground image and the background image are identified along a boundary associated with a location at which to paste the foreground image relative to the background image. Thereafter, recursive down sampling of a region of pixels within the boundary by a sampling factor is performed to produce a plurality of down sampled images having color difference indicators associated with each pixel of the down sampled images. Such color difference indicators indicate whether a difference of color value exists for the corresponding pixel. To effectuate a seamless transition, the color difference indicators are normalized in association with each recursively down sampled image.

Abstract: Systems and methods are provided for in-context editing of websites and other output presentations via automatic pattern detection. One embodiment involves an editor application receiving an output presentation. The output presentation is generated based on at least one source element from at least one source file. The editor application identifies at least one pattern among multiple output elements in the output presentation. The at least one pattern corresponds to the at least one source element. The editor application determines a modification to the source element corresponding to a modification to at least one of the output elements. The modification to the source element is determined based on identifying the at least one pattern. The editor application applies the corresponding modification to the at least one source element responsive to receiving the modification to the at least one output elements.

Abstract: Various aspects of this disclosure describe setting an audio compressor threshold using averaged audio measurements. Examples include calculating one or more average values of amplitude values of an audio file, and setting a threshold used in the audio compressor based on the calculated thresholds. Samples of the audio file with amplitude values above the threshold are attenuated, while samples of the audio file with amplitude values below the threshold are not attenuated. The threshold can be set equal to a calculated average value, or from a function of one or more calculated average values. Different audio channels comprising the audio file can be processed to set a respective compressor threshold for each audio channel.

Abstract: A digital medium environment includes an action processing application that performs actions including personalized recommendation. A learning algorithm operates on a sample-by-sample basis (e.g., each instance a user visits a web page) and recommends an optimistic action, such as an action found by maximizing an expected reward, or a base action, such as an action from a baseline policy with known expected reward, subject to a safety constraint. The safety constraint requires that the expected performance of playing optimistic actions is at least as good as a predetermined percentage of the known performance of playing base actions. Thus, the learning algorithm is conservative during exploratory early stages of learning, and does not play unsafe actions. Furthermore, since the learning algorithm is online and can learn with each sample, it converges quickly and is able to track time varying parameters better than learning algorithms that learn on a block basis.

Abstract: An authorized user obtains a packaging license that grants permission to use a particular recording device to generate multimedia content in accordance with specified license terms. The packaging license includes a content key that is used to encrypt the multimedia content at the point of capture on the recording device. The encrypted multimedia content can be transmitted via unsecure channels (for example, via electronic mail) to a networked content repository or an intended recipient. For playback, an authorized user obtains a playback license that grants permission to decrypt and playback the multimedia content using a particular playback device. An authorization server and a key management server are used to manage which users are entitled to receive a license, and to define the terms of the granted licenses. A record of the granted authorizations and licenses is maintained, thereby allowing access to a given content item to be audited.