Microsoft Technology Licensing, LLC

Other

Microsoft Corporation

Ross Garrett Cutler

Yong Rui

Anoop Gupta

Senthil Kumar Velayutham

Zhengyou Zhang

Zicheng Liu

Ramin Mehran

Anton W Krantz

Cha Zhang

Andrew Peter Sinclair

Martin Ellis

Jonathan Adam Kauffman

Xinding Sun

Jaysun Stockdell-Giesler

Joyce S Fang

Henrique Sarmento Malvar

Paul A Viola

Li-wei He

Ankit Tandon

Antti Kelloniemi

Pei Yin

Johnathan J Cadiz

Christian F Huitema

Michael F Cohen

Ivan Jelev Tashev

Sriram Srinivasan

David Yuheng Zhao

Stanley Thomas Birchfield

Daniel Allen Rosenfeld

Avronil Bhattacharjee

Adam G Kirk

Oliver Arthur Whyte

Thang Duong

Sailaja Malladi

Robert Aichner

Adarsh Prakash Murthy Kowdle

Karlton David Sequeira

Vishak Gopal

Jayant Ajit Gupchup

Ju Hee Cho

Tommi Antero Raussi

Alejandro Acero

John R Douceur

Philip Andrew Chou

Timothy Mark Moore

Jason Garms

Kentaro Toyama

Dawson Yee

Daniel Chaim Robbins

Dinei A Florencio

Matthew Uyttendaele

Ashish Kapoor

Simon Clarke

Jeremy Thomas Buch

Scott C Forbes

Nuria M Oliver

William George Verthein

Arash Ghanaie-Sichanie

Roy A Leban

Kori Inkpen Quinn

Eric K Ringger

William Buxton

Kori Marie Quinn

Simone Leorin

Ramaswamy Hariharan

Josh Watson

Daniel J Sisolak

Alan L Bridgewater

Armen Hovhannisyan

Anthony J Bawcutt

Chandan Karadagur Ananda Reddy

Harishchandra Dubey

Phillip A Chou

Ya Chang

John E Unverferth, Iii

Senthil K Velayutham

Antti Pekka Kelloniemi

Senthil Velayutham

This document relates to active speaker detection using distributed devices. For example, the disclosed implementations can employ personal devices of one or more users to detect when those users are speaking during a call with other users. Then, a camera on the personal device can be employed to obtain a front-facing view of the user, which can be provided to other call participants. In some cases, a microphone and/or camera on the user's device are employed to detect when the user is actively speaking.

Active speaker detection using distributed devices

This document relates to distributed teleconferencing. Some implementations can employ personalized enhancement models to enhance microphone signals for participants in a call. Further implementations can perform proximity-based mixing, where microphone signals received from devices in a particular room can be omitted from playback signals transmitted to other devices in the same room. These techniques can allow enhanced call quality for teleconferencing sessions where co-located users can employ their own devices to participate in a call with other users.

Distributed teleconferencing using personalized enhancement models

This document relates to distributed teleconferencing. Some implementations can employ adaptive audio or video enhancement to address scenarios where audio enhancement can tend to remove desirable sounds. For instance, adaptive audio enhancement can involve detecting the presence of a sound, such as clapping, and modifying audio enhancement so that the sound is retained in an enhanced audio signal. Adaptive video processing can involve detecting the presence of the sound and adding a graphical identifier to a video signal that conveys the presence of that sound.

Adaptive enhancement of audio or video signals

This document relates to training and employing of quality estimation models to estimate the quality of different signal characteristics. One example includes a method or technique that can be performed on a computing device. The method or technique can include obtaining training signals exhibiting diverse impairments introduced when the training signals are captured or diverse artifacts introduced by different processing characteristics of a plurality of data enhancement models. The method or technique can also include obtaining quality labels for different signal characteristics of the training signals. The method or technique can also include training at least two different quality estimation models to estimate quality of at least two different signal characteristics based at least on the training signals and the quality labels.

Quality estimation models for various signal characteristics

An agent interfaces with a sending computing device and a receiving computing device to automatically adjust one-way or two-way real-time audio and real-time video transmission parameters responsive to changing network conditions and/or application requirements. The agent incorporates a reinforcement learning model that adjusts transmission parameters to maximize an expected value of a sum of future rewards; the expected value of the sum of future rewards is based on a current state of the sending computing, a current action (e.g. a current set of transmission parameters) at the sending computing device and a reward provided by the receiving computing device. The reward is representative of a user-perceived quality of experience at the receiving computing device.

Reinforcement learning in real-time communications

A system is provided herein that can correlate service issues with system telemetry associated with the software session associated with those service issues. Using a statistical approach, the system can evaluate data across numerous software sessions to rank the importance of the reported service issues. To accomplish the ranking, the system can parse the reports of service issues on a periodic basis, can extract telemetry identifiers (IDs) from the logs, can query the telemetry, may compute the relative importance of detected issues (in the context of calls going on for that day), and then can report this impact hack to the service issue database.

Service issue prioritisation based on impact using software telemetry

Techniques in connection with a light field camera array are disclosed, involving generating a temperature data for an imaging camera included in an imaging camera array for a first time, obtaining an image data from the imaging camera, generating temperature-based correction parameters corresponding to the temperature data based on at least a stored temperature calibration data; and producing corrected image data by applying a geometric distortion correction and/or color correction indicated by the temperature-based correction parameters to the image data.

Dynamic detection and correction of light field camera array miscalibration

User-input obfuscating techniques involving displaying multiple targets on a display device. A visual indication of the current user target moves to a first target at a first time in response to user input, and a first presentation of a visual or audio feedback signal confirming actuation of a target occurs at a second time, without a user input for actuation of a target received while the current user target has remained at the first target from the first time to the second time. The visual indication moves to a second target at a third time in response to user input. The second target is actuated in response to receiving user input for actuation of the current user target after the third time. Then, with the current user target having remained at the second target since the third time, a second presentation of the visual or audio feedback signal is provided.

Systems and methods of selection acknowledgement for interfaces promoting obfuscation of user operations

A camera system including a first imaging sensor having a first imaging surface with a first diagonal length, a first lens arranged to guide a first image formation light flux toward the first imaging surface with the first image formation light flux having at the first imaging surface a width equal to or greater than the first diagonal length, a second imaging sensor having a second imaging surface with a second diagonal length, a second lens arranged to guide a second image formation light flux toward the second imaging surface with the second image formation light flux having at the second imaging surface a width equal to or greater than the second diagonal length. The first lens and the second lens are oriented in opposing directions, and the first imaging sensor, the first lens, the second imaging sensor, and the second lens are each mounted partially within an enclosure.

Panoramic camera system

Techniques for obfuscating user selections during user-input of sensitive information, involving receiving a user input entered via a user input device and, in response, causing a first visual indication to move over a period of time through a first subset of displayed targets to enter user selections of targets. In a concurrent process, control signals are generated to display a second visual indication in addition to the first visual indication. The second visual indication moves through a second subset of the displayed targets during the period of time in a manner that is different from the first visual indication moving through the first subset of the displayed targets. The current user target is thus obfuscated as two or more different visual indicators move through the displayed targets.

Systems and methods for obfuscating user selections

Techniques in connection with a light field camera array are disclosed, involving obtaining a first image data from a first image camera included in the light field camera array for a first time, obtaining a second image data from the first image camera included in the light field camera array for a second time before the first time, obtaining a third image data from a second image camera included in the light field camera array at a different position than the first image camera and having a field of view overlapping a field of view of the first imaging camera, detecting an inconsistency in a view of a scene between the first image data and the second image data and/or the third image data, automatically attempting to generate, in response to the detection of the inconsistency, correction data for the first image camera to reduce or eliminate the detected inconsistency.

Techniques for obfuscating selection of an option during user-input of sensitive information, involving displaying several targets, such as via a virtual keyboard, on a device display, where each target corresponds to one of multiple options. When a first visual indication of a current user target moves to a first target in response to a first user navigation input, an obfuscation system can concurrently or soon thereafter move a second visual indication of a first distractor target to another target. The current user target is thus obfuscated as two different visual indicators move through the targets.

Systems and methods for input interfaces promoting obfuscation of user navigation and selections

A system is provided herein that can correlate service issues with system telemetry associated with the software session associated with those service issues. Using a statistical approach, the system can evaluate data across numerous software sessions to rank the importance of the reported service issues. To accomplish the ranking, the system can parse the reports of service issues on a periodic basis, can extract telemetry identifiers (IDs) from the logs, can query the telemetry, may compute the relative importance of detected issues (in the context of calls going on for that day), and then can report this impact back to the service issue database.

Techniques for obfuscating user selections during user-input of sensitive information, involving displaying targets, such as keys of a virtual keyboard, on a display device, where each target corresponds to one of multiple options. A visual indication for a current user target can move to a first target in response to a first input event at a first time, and the first target can be actuated in response to a second input event for actuation of the current user target at a second time. An obfuscation system is configured to present to a user, at approximately the second time, a discriminatory signal used to confirm actuation of a target. The discriminatory signal includes one of a haptic signal generated by a device in physical contact with the user, an audio signal, and a modified image.

Growing community of inventors

Duvall, WA, United States of America

Ross Garrett Cutler