Techniques for time-correlated ink are described. According to various embodiments, ink input is correlated to content. For instance, ink input received during playback of a video is timestamped. According to various embodiments, ink input displayed over content is removed after input ceases. Further, ink input is displayed during playback of the portion of content to which the ink input is time correlated.

Described is a solution for an unlabeled target domain dataset challenge using a domain adaptation technique to train a neural network using an iterative 3D model fitting algorithm to generate refined target domain labels. The neural network supports the convergence of the 3D model fitting algorithm and the 3D model fitting algorithm provides refined labels that are used for training of the neural network. During real-time inference, only the trained neural network is required. A convolutional neural network (CNN) is trained using labeled synthetic frames (source domain) with unlabeled real depth frames (target domain). The CNN initializes an offline iterative 3D model fitting algorithm capable of accurately labeling the hand pose in real depth frames. The labeled real depth frames are used to continue training the CNN thereby improving accuracy beyond that achievable by using only unlabeled real depth frames for domain adaptation.

In various example embodiments, a system and method for determining an item that has confirmed characteristics are described herein. An image that depicts an object is received from a client device. Structured data that corresponds to characteristics of one or more items are retrieved. A set of characteristics is determined, the set of characteristics being predicted to match with the object. An interface that includes a request for confirmation of the set of characteristics is generated. The interface is displayed on the client device. Confirmation that at least one characteristic from the set of characteristics matches with the object depicted in the image is received from the client device.

Embodiments provide for perceptual data association from at least a first and a second sensor disposed at different positions in an environment, in respective series of local scene graphs that identify characteristics of objects in the environment that are updated asynchronously and merging the series of local scene graphs to form a coherent image of the environment from multiple perspectives.

Embodiments provide for perceptual data association from at least a first and a second sensor disposed at different positions in an environment, in respective series of local scene graphs that identify characteristics of objects in the environment that are updated asynchronously and merging the series of local scene graphs to form a coherent image of the environment from multiple perspectives.

A pose determination method for a subject includes identifying a plurality of candidate regions from depth data representing an environment based on a depth connectivity criterion, determining a first region including a first subset of the plurality of candidate regions based on an estimation regarding a first pose component of the subject, determining a second region including a second subset of the plurality of candidate regions that are disconnected from the first subset of the plurality of candidate regions based on relative locations of the first region and the second region, generating a collective region by associating the first region with the second region, identifying the first pose component and a second pose component of the subject from the collective region, determining a spatial relationship between the first pose component and the second pose component, and generating a controlling command based on the spatial relationship.

A method for classifying subject activities in videos includes learning latent (previously generated) concepts that are analogous to nodes of a graph to be generated for an activity in a video. The method also includes receiving video segments of the video. A similarity between the video segments and the previously generated concepts is measured to obtain segment representations as a weighted set of latent concepts. The method further includes determining a relationship between the segment representations and their transitioning pattern over time to determine a reduced set of nodes and/or edges for the graph. The graph of the activity in the video represented by the video segments is generated based on the reduced set of nodes and/or edges. The nodes of the graph are represented by the latent concepts. Subject activities in the video are classified based on the graph.

Various embodiments describe techniques for making inferences from graph-structured data using graph convolutional networks (GCNs). The GCNs use various pre-defined motifs to filter and select adjacent nodes for graph convolution at individual nodes, rather than merely using edge-defined immediate-neighbor adjacency for information integration at each node. In certain embodiments, the graph convolutional networks use attention mechanisms to select a motif from multiple motifs and select a step size for each respective node in a graph, in order to capture information from the most relevant neighborhood of the respective node.

Graph-theoretic segmentation methods for segmenting histological structures in H&E stained images of tissues. The method rely on characterizing local spatial statistics in the images. Also, a method for quantifying intratumor spatial heterogeneity that can work with single biomarker, multiplexed, or hyperplexed immunofluorescence (IF) data. The method is holistic in its approach, using both the expression and spatial information of an entire tumor tissue section and/or spot in a TMA to characterize spatial associations. The method generates a two-dimensional heterogeneity map to explicitly elucidate spatial associations of both major and minor sub-populations.

Systems and methods are disclosed to receive an image depicting at least a part of a document and identify a plurality of partition points dividing the image into potential segments; generate a linear partition graph (LPG) comprising a plurality of vertices using the plurality of partition points and a plurality of arcs connecting the plurality of vertices; identify a path of the LPG having a value of a quality metric above a threshold value, wherein the path is selected from a plurality of paths of the LPG and comprises one or more arcs and the value of the quality metric is derived using a neural network classifying each of a plurality of pixels of the image; and generate one or more blocks of the image wherein each of the one or more blocks corresponds to an arc of the identified path and represents a portion of the image associated with a type of an object.