Abstract: Techniques for indoor location-finding are described herein. Inside buildings, anomalies in the Earth's magnetic field may be caused by the buildings' structures. These anomalies tend to be location specific and temporally stable. Accordingly, a database or map may be constructed of field strength measurements, and made available to clients, either online (e.g., through cloud services) or offline. As a user carrying a portable device walks within a building, magnetic field measurements may be made. The measurements may be configured in as vectors and compared to the database. A specific location may be indicated by anomaly information seen in both the vector(s) and database. Optionally, techniques may be combined with a particle filter that may be operated step-by-step as the user moves within the building. In another option, fusion with Wi-Fi signals may be performed, such as with a two-pass bidirectional particle filtering.

Abstract: Step detection and step length estimation techniques may include detecting salient points in sensor data of one or more sensors. A step frequency may be used to estimate the length of a step according to a step length estimation model. The step length estimation model may be adjusted based at least in part on landmark data to better estimate a step length of the user. Additionally or alternatively, an adjusted step length estimation model may be readjusted over time to account for changes in a user, conditions, or both.

Abstract: Step detection and step length estimation techniques include detecting salient points in sensor data of one or more sensors. A step frequency is estimated based on a time interval between the detected salient points. A step length of the step may then be computed based on a nonlinear combination of the estimated step frequency and a function of the sensor data, and/or a step model. Alternatively, the step length of the step may be computed based on a combination of a nonlinear function of the estimated step frequency and a (linear or nonlinear) function of the sensor data, and/or a step model.

Abstract: A user device is equipped with one or more sensors that collect data relating to a movement of a user having the user device. An estimated area within which a starting location of the user device is located is determined based on one or more signals received from adjacent signal sources. A plurality of candidate locations is identified within the estimated area as the potential starting location of the user device. Map information of surrounding area that covers the estimated area is also obtained. One or more candidate locations are filtered out depending on whether they could have experienced the movement based on the map information. When one or more re-starting conditions have met during the movement, a then starting location of the user device needs to be determined.

Abstract: Technologies pertaining to computing depth images of a scene that includes a mobile object based upon the principle of light falloff are described herein. An infrared image of a scene that includes a mobile object is captured, wherein the infrared image has a plurality of pixels having a respective plurality of intensity values. A depth image for the scene is computed based at least in part upon square roots of respective intensity values in the infrared image.

Abstract: Techniques for utilizing an infrared illuminator, an infrared camera, and a projector to create a virtual 3D model of a real 3D object in real time for users' interaction with the real 3D object.

Abstract: Technology is described for determining a location at which a computing device is positioned. For example, a computing device is positioned in an area (e.g., building), and a map (e.g., floor plan) is retrieved that depicts the area. An initial location of the computing device is determined with respect to the map. Inertial sensors record motion inputs (e.g., acceleration, orientations, etc.), which are analyzed to determine a path along which the computing device moves. The path is applied to the initial location to determine an updated location at which the computing device may be located.

Abstract: Step detection and step length estimation techniques include detecting salient points in sensor data of one or more sensors. A step frequency is estimated based on a time interval between the detected salient points. A step length of the step may then be computed based on a nonlinear combination of the estimated step frequency and a function of the sensor data, and/or a step model. Alternatively, the step length of the step may be computed based on a combination of a nonlinear function of the estimated step frequency and a (linear or nonlinear) function of the sensor data, and/or a step model.

Abstract: Methods, computer systems, and computer-readable media for accelerating a learning-to-rank algorithm using a central processing unit (CPU) and a graphics processing unit (GPU) are provided. The GPU processes document pairs created by the CPU in parallel to generate a lambda-gradient value and a weight for each document. The CPU builds a regression tree for the documents. The GPU is utilized to accelerate this process by constructing histograms of feature values, wherein the address of bins collecting the same feature value are shifted during the construction of the histogram. The output of the regression tree is a score for each document which is used to rank or order the document on a search engine results page.

Abstract: In some implementations, an access device may sense movement in six degrees of freedom for interacting with information in a user interface. In some implementations, a security feature may be included for verifying an authorization of a user to use the access device. Additionally, in some implementations, the access device may establish communication with an available computing device for accessing and interacting with information that may include remotely stored information, cloud data, cloud services, and the like.

Abstract: Techniques for utilizing an infrared illuminator, an infrared camera, and a projector to create a virtual 3D model of a real 3D object in real time for users' interaction with the real 3D object.

Abstract: Techniques for utilizing two or more mobile devices equipped with projectors to generate a combined seamless user interfaces by stitching projection areas generated by the projectors.