Microsoft Technology Licensing, LLC

Hewlett-packard Development Company, L.p.

Amazon Technologies, Inc.

Adobe, Inc.

Lior Shapira

Ran Gal

Eyal Ofek

Douglas Christopher Burger

Pushmeet Kohli

Judith Amores Fernandez

Xavier Benavides Palos

Andrew Jay Roths

Jiahe Helen Wang

Carl Staelin

Renato Keshet

Pavel Kisilev

Shmuel Avidan

Nathan Silberman

Benjamin Livshits

Michal Aharon

David Alexander Molnar

Gregory Braverman

Alexander Nikolavich Moshchuk

Daniel Freedman

Boris Oicherman

Zachi Karni

Kadirvel Chockalingam Vanniarajan

Yohanan Sivan

Omar Abdul Baki

Alexander Moshchuk

Blair MacIntyre

Sudharsan Sampath

John Vilk

Blair Macintyre

A device provisioning service (DPS) fields requests from unprovisioned devices so that those unprovisioned devices can obtain network credentials or other data used in provisioning the unprovisioned device. The DPS can identify the device securely and associate with a known user account, or the user provisioning the device can supply network credentials over a side channel after supplying a provision code indicative of possession of the unprovisioned device. The provision code can be unique to the unprovisioned device or a short-sequence code that is not necessarily unique, but that is sufficiently uncommon that a specific short-sequence code would not likely be used more than once at a time. In order to communicate with the DPS, a provisioning device might connect the unprovisioned device and the DPS. If the provisioning device is a trusted device, it can perform some of the steps otherwise required by the DPS.

Simplified association of devices with a network using unique codes on the devices and side channel communication

Tactile virtual reality (VR) and/or mixed reality (MR) experiences are described. Techniques described herein include receiving data from a sensor and accessing a position and an orientation of a real object that is physically present in a real scene. Furthermore, techniques described herein include identifying the real object based at least in part on the position and the orientation of the real object and causing a graphical element corresponding to the real object to be rendered on a display of a VR and/or MR display device. The graphical element can be determined based at least in part on a VR and/or MR application. The techniques described herein include determining an interaction with the real object and causing a functionality associated with the graphical element to be performed in the VR or MR environment rendered via the VR and/or MR display device, respectively.

Tactile interaction in virtual environments

A 'Layout Optimizer' provides various real-time iterative constraint-satisfaction methodologies that use constraint-based frameworks to generate optimized layouts that map or embed virtual objects into environments. The term environment refers to combinations of environmental characteristics, including, but not limited to, 2D or 3D scene geometry or layout, scene colors, patterns, and/or textures, scene illumination, scene heat sources, fixed or moving people, objects or fluids, etc., any of which may evolve or change over time. A set of parameters are specified or selected for each object. Further, the environmental characteristics are determined automatically or specified by users. Relationships between objects and/or the environment derived from constraints associated with objects and the environment are then used to iteratively determine optimized self-consistent and scene-consistent object layouts. This enables the Layout Optimizer to augment environments with arbitrary content in a structured constraint-based process that adapts to changing scenes or environments.

Layout design using locally satisfiable proposals

An 'Anchored Environment Generator' generates a physically constrained virtual environment that is molded and anchored to a real-world environment around a user (or multiple users). This molding and anchoring of the physically constrained virtual environment ensures that at least a portion of the physically constrained virtual environment matches tactile truth for one or more surfaces and objects within the real-world environment. Real objects and surfaces in the real-world environment may appear as different virtual objects, and may have different functionality, in the physically constrained virtual environment. Consequently, users may move around within the physically constrained virtual environment while touching and interacting with virtual objects in the physically constrained virtual environment. In some implementations, the physically constrained virtual environment is constructed from virtual building blocks that are consistent with a theme-based specification (e.g., French bistro, space station, starship, alternate real-world environments, alien world, apocalyptic wasteland, medieval village, hospital, castle, etc.).

Molding and anchoring physically constrained virtual environments to real-world environments

A 'Shared Tactile Immersive Virtual Environment Generator' (STIVE Generator) constructs fully immersive shared virtual reality (VR) environments wherein multiple users share tactile interactions via virtual elements that are mapped and rendered to real objects that can be touched and manipulated by multiple users. Generation of real-time environmental models of shared real-world spaces enables mapping of virtual interactive elements to real objects combined with multi-viewpoint presentation of the immersive VR environment to multiple users. Real-time environmental models classify geometry, positions, and motions of real-world surfaces and objects. Further, a unified real-time tracking model comprising position, orientation, skeleton models and hand models is generated for each user. The STIVE Generator then renders frames of the shared immersive virtual reality corresponding to a real-time field of view of each particular user. Each of these frames is jointly constrained by both the real-time environmental model and the unified real-time tracking model.

Shared tactile interaction and user safety in shared space multi-person immersive virtual reality

A 'Tactile Autonomous Drone' (TAD) (e.g., flying drones, mobile robots, etc.) supplies real-time tactile feedback to users immersed in virtual reality (VR) environments. TADs are not rendered into the VR environment, and are therefore not visible to users immersed in the VR environment. In various implementations, one or more TADs track users as they move through a real-world space while immersed in the VR environment. One or more TADs apply tracking information to autonomously position themselves, or one or more physical surfaces or objects carried by the TADs, in a way that enables physical contact between those surfaces or objects and one or more portions of the user's body. Further, this positioning of surfaces or objects corresponds to some real-time virtual event, virtual object, virtual character, virtual avatar of another user, etc., in the VR environment to provide real-time tactile feedback to users immersed in the VR environment.

Autonomous drones for tactile feedback in immersive virtual reality

An 'AR Privacy API' provides an API that allows applications and web browsers to use various content rendering abstractions to protect user privacy in a wide range of web-based immersive augmented reality (AR) scenarios. The AR Privacy API extends the traditional concept of 'web pages' to immersive &#x201c;web rooms&#x201d; wherein any desired combination of existing or new 2D and 3D content is rendered within a user's room or other space. Advantageously, the AR Privacy API and associated rendering abstractions are useable by a wide variety of applications and web content for enhancing the user's room or other space with web-based immersive AR content. Further, the AR Privacy API is implemented using any existing or new web page coding platform, including, but not limited to HTML, XML, CSS, JavaScript, etc., thereby enabling existing web content and coding techniques to be smoothly integrated into a wide range of web room AR scenarios.

Protecting privacy in web-based immersive augmented reality

Surface reconstruction contour completion embodiments are described which provide dense reconstruction of a scene from images captured from one or more viewpoints. Both a room layout and the full extent of partially occluded objects in a room can be inferred using a Contour Completion Random Field model to augment a reconstruction volume. The augmented reconstruction volume can then be used by any surface reconstruction pipeline to show previously occluded objects and surfaces.

Contour completion for augmenting surface reconstructions

According to an example, data pertaining to a plurality of entities recommended for a user may be accessed, in which the data identifies relationships between the plurality of entities with respect to each other. In addition, a relevance map for the user that displays graphical representations of the plurality of entities over a substantially optimized use of space available for display of the graphical representations in the relevance map may be generated, in which the graphical representations of the plurality of entities are arranged in the relevance map according to a predetermined arrangement scheme.

Displaying recommended entities in a relevance map

A method comprises computing an activity map for an input image. The activity map indicates an amount of variable spatial activity in the input image. The method further comprises producing an output image based on the activity map, and controlling a printhead to cause the output image to be printed. The output image contains a plurality of color pixels and each such color pixel is represented by at least one of cyan (C), magenta (M), yellow (Y), and black (K) in a color space.

Determining new color values of an image based on an activity map

A clustering method for high-dimensionality data includes identifying a set of nearest neighbors of a point in a multidimensional space and determining the centroid of the set of nearest neighbors, where the centroid is a member of the set of nearest neighbors. The method is then repeated using the neighbors identified around the computed centroid. In one embodiment, the method may terminate when the computed centroid becomes stationary over successive iterations. The resulting centroid may be returned as a mode of the data set. Points of the data set having common modes may be assigned to the same cluster.

Clustering techniques for large, high-dimensionality data sets

Growing community of inventors

Redmond, WA, United States of America

Lior Shapira