Microsoft Technology Licensing, LLC

Other

Amazon Technologies, Inc.

Andrew Stuart Glass

Christian Klein

Sang Ho Yoon

Roland Radtke

Wade H Curtiss

Margaret Hope Magnus

Adam James Betz

Yongjun Wu

Jan-Kristian Markiewicz

Lei Li

Claes-Fredrik Urban Mannby

Stefan Christian Richter

Amarsingh Buckthasingh Winston

Erik Fortune

Edward Helbling

Nicholas James Benson

Srikanth Kiran Kotagiri

Andrew D Rosen

Rylan Michael Hawkins

Peter Geoffrey Constable

Eric Scott Albright

Corey Joseph Loman

Joseph Bernard Tobens, Iii

Rakesh Tangirala

Mark Rodger Rideout

Roy H Berger

Samuel Aaron Beard

Microsoft Technology Licensing

Gyancarlo García Avilla

Ana Isabel A N Ribeiro Alberts

Daniel Brunner

Nick Beal

In non-limiting examples of the present disclosure, systems, methods and devices for filtering predictive text surfacing candidates are provided. A predictive text filtering model may be maintained. The predictive text filtering model may comprise a plurality of terms that are associated in the predictive text filtering model with a precarious classification, and a range of a number of words for filtering cooccurrences of precarious and other precarious or blocklist terms from the plurality of terms. A text input may be processed with a predictive text model. A plurality of surfacing candidates may be determined based on the processing of the text input with the predictive text model. The predictive text filtering model may be applied to a surfacing candidate of the plurality of candidates. The surfacing candidate may be filtered from further processing. At least one non-filtered surfacing candidate may be displayed.

Post-model filtering of predictive text

In non-limiting examples of the present disclosure, systems, methods and devices for determining a language of a text string are presented. A language detection model may be maintained. The language detection model may comprise identities and weights for initial and final consonants, identities and weights for prefixes and suffixes, and identities and weights for vowel sequences, where each identity is derived from a training corpus. The weights may correspond to a frequency of a text unit in the corpus. A text string may be received and a match score between the text string and the language of the language detection model may be determined. The match score may be based on initial and final consonant scores, prefix and suffix scores, and/or vowel sequence scores for each word in the text string. If the match score meets a threshold value a follow-up action associated with the language may be performed.

Training a language detection model for language autodetection from non-character sub-token signals

The disclosed technologies address various technical and user experience problems by augmenting the functionality provided by virtual on-screen trackpads without requiring users to switch between modes. In this way, users can access extended functionality without interfering with expected traditional trackpad functionality (e.g. moving a cursor, clicking, and scrolling). In particular, technologies are disclosed for modeless digital pen input to a virtual trackpad, modeless gestures for summoning user interfaces, modeless gestures utilizing position relative to adjacent user interfaces, and modeless cursor control and interaction with virtual touch targets.

Modeless augmentations to a virtual trackpad on a multiple screen computing device

The disclosed technologies address various technical problems, including improving human-computer interaction, by augmenting the functionality provided by user input devices using a digital glove. For example, a digital glove can be utilized to provide pressure sensitivity for non-pressure sensitive touchscreen devices, attribute input made with a digital pen to a user wearing the digital glove, augment the functionality provided by mice and other types of handheld user input devices, augment the functionality provided gaze tracking systems, improve camera-based motion tracking systems, and to augment the functionality provided by speech recognition systems.

Augmenting the functionality of user input devices using a digital glove

Language autodetection from non-character sub-token signals

The disclosed technologies address technical problems, including improving human-computer interaction, by augmenting the functionality provided by non-digital objects using a digital glove. To provide this functionality, a machine learning model is trained using sensor data generated by sensors in a digital glove and data generated by a user input device while the digital glove is utilized to manipulate an object like a user input device. Once trained, the machine learning model can take sensor data generated by a digital glove while manipulating a non-digital object and generate virtual user input device data that can be utilized to control a host computer. A digital glove can also be utilized to perform selection operations using non-digital objects when pressure data generated by one or more of the pressure sensors in the digital glove indicates that pressure was exerted at a finger of the digital glove in excess of a threshold value.

Augmenting the functionality of non-digital objects using a digital glove

Techniques for bi-manual control of an operation may allow gestural input for the operation to be inputted blindly with one hand while a user's attention remains on the other hand. An operation invoked and controlled by the gestural input may automatically associate with an operation controlled by the other hand. The gestural input may facilitate blind control of the operation by being able to be initially recognized without regard for the scale, orientation, and/or location of the invoking gestural input. Moreover, scale and/or orientation of the operation may be determined by the gestural input. Changes in scale and/or orientation of the gestural input during the blind operation may change the scale and/or orientation of the operation. The operation may manifest by displaying corresponding graphics at a location independent of the gestural input or by associating functionality of the blind operation with an operation performed by the other hand.

User interfaces for bi-manual control

A device may be configured to enable a user to select a language, and may fulfill resource requests from applications by selecting, from among resources respectively associated with a language, a resource associated with the selected language of the user. However, this resource selection process may be inadequate if the user selects multiple languages; if a resource associated with the selected language of the user is unavailable, but resources associated with related languages are available; or if the user and/or the application specifies an ordering for the selection among the languages. Presented herein are techniques for performing the resource selection by, for respective languages selected by the user, calculating a weight representing a suitability of the language for the resource request; generating a selection order of the selected languages according to the weights; and selecting a resource based on the position of the associated language in the selection order.

Language lists for resource selection based on language text direction

Techniques are described for server-side fragment insertion and delivery of media content. After receiving a request to play back media content, a manifest is generated that includes fragments associated with the requested media content and fragments associated with different media content. A client device requests fragments from a media server according to the manifest, and the media server provides the correct fragment in response to the request.

Server-side fragment insertion and delivery

The techniques disclosed herein enable dynamic device interaction adaptation based on user engagement. In general, a computer can leverage sensors to determine a user's level of engagement with one or more devices. As the user transitions his or her attention to different devices, the computer can utilize different interaction models to assist the user in interacting with each device that meets a threshold level of engagement with the user. Individual interaction models can configure the computer to direct input gestures received by the computing device to a selected device, and adapt the computer's input modality to provide user interface controls that are optimized for the selected device. When a user is interacting with two or more devices, automatic transitions between different interaction models can help improve the user's interaction with each device.

Dynamic device interaction adaptation based on user engagement

A device may provide a login process to authenticate users prior to admittance to a computing environment. The device may also enable users to adjust various the computing environment, e.g., the language selected for communicating with the user and the user interfaces to be presented to the user, and may store such adjustments in a secured user account. However, if the user account is inaccessible to the device during the login process, the device is unable to adapt the login process to apply the user's adjustments. Instead, the device may be configured to store users' adjustments (including language selection) outside of the user accounts, and to, upon identifying the user during the login process, present login interfaces specified in the user account. Additionally, users may select different login interfaces during login, and the device may retrieve these login interfaces for selection during future login processes for the same user.

Login interface selection for computing environment user login

Growing community of inventors

Seattle, WA, United States of America

Andrew Stuart Glass