Abstract: A system, method and computer-readable storage device provides an improved speech processing approach in which hyper parameters used for speech recognition are modified dynamically or in batch mode rather than fixed statically. The method includes estimating, via a model trained on audio data and/or metadata, a set of parameters useful for performing automatic speech recognition, receiving speech at an automatic speech recognition system, applying, by the automatic speech recognition system, the set of parameters to processing the speech to yield text and outputting the text from the automatic speech recognition system.

Abstract: Improved compilers recognize interception declarations in source code, and emit code that replaces invocations of specified artifacts with invocations of specified interceptors instead. Source generators proactively modify program behavior in arbitrary desired ways without introducing hidden security violations and without requiring edits by the program's developers. Interception declarations are visible in source code and development tools. In some cases, different invocations of a particular method at respective locations are intercepted by different replacement methods. Replacement methods have identical signatures, or are otherwise compatible. Some interceptors specify optional parameters. Method calls, field accesses, and property calls may be intercepted. Work to modify program behavior is moved in the program lifecycle from runtime to compile time, thus improving runtime performance and eliminating JIT compilation security risks.

Abstract: The disclosure derives insight from telemetry data by receiving telemetry data; parsing the received telemetry data to identify properties, and mapping the identified properties to a set of identified tags based at least on a tag library. Based at least on the mapping, the disclosure generates insight data and a report for the telemetry data. In this manner, the disclosure adds structure to data, thereby providing semantic meaning to internet of things (IoT) telemetry data, regardless of the device class or manufacturer. This, in turn, automatically creates applicable insights. Insights are available with a core tag taxonomy, which can be extended and customized. By applying the tags, a user obtains immediate insights related to usage, performance, and health of a monitored product or service.

Abstract: A mixed semiconductor-superconductor platform is fabricated in phases. In a masking phase, a dielectric mask is formed on a substrate, such that the dielectric mask leaves one or more regions of the substrate exposed. In a selective area growth phase, a semiconductor material is selectively grown on the substrate in the one or more exposed regions. In a superconductor growth phase, a layer of superconducting material is formed, at least part of which is in direct contact with the selectively grown semiconductor material. The mixed semiconductor-superconductor platform comprises the selectively grown semiconductor material and the superconducting material in direct contact with the selectively grown semiconductor material.

Abstract: A gaze detection assembly comprises a plurality of infrared (IR) light emitters configured to emit IR light toward a user eye and an IR camera configured to sequentially capture IR images of the user eye. A controller is configured to, during a reference frame, control the plurality of IR light emitters to emit IR light toward the user eye with a reference intensity distribution. During the reference frame, an IR image is captured depicting a first glint distribution on the user eye. Based at least in part on the first glint distribution, during a subsequent frame, the plurality of IR light emitters are controlled to emit IR light toward the user eye with a subsequent intensity distribution, different than the reference intensity distribution. During the subsequent frame, a second IR image is captured depicting a second glint distribution on the user eye.

Abstract: Technology is disclosed for controlling the processing of electronic communications on computing devices. An electronic communication is processed to determine mentions in the message body indicating recipients of the communication. A point-of-view (POV) is determined for each mention, with respect to recipient(s), as second-person or third-person POV. The communication is parsed into sections, and the mentions that are associated with each section are determined. Based on the POV for a mention associated with a section, it is determined that the section is directed to the recipient(s) indicated by the mention, if the POV of the mention is second-person POV, or the section is relevant to the recipient(s) indicated by the mention, if the POV of the mention is third-person POV. Additionally, an enhanced communication data is generated indicating the sections and corresponding POVs of mentions associated with the sections, and used to provide a personalized computing experience to users.

Abstract: In non-limiting examples of the present disclosure, a user input device is provided. The user-input device may include non-contact magnetic, capacitive, or inductive components for translating movements of a joystick into angular positions. Magnets attached to axial shafts of a joystick assembly may cause changes in voltage in magnetometers on a PCB when the axial shafts are rotated, which may be translated into angular positions of the joystick. Conductors attached to axial shafts of a joystick assembly may cause changes in capacitive charge in capacitors on a PCB when the axial shafts are rotated, which may be translated into angular positions of the joystick. Anisotropic magnetically permeable material and/or excitation components attached to axial shafts of a joystick assembly may cause changes in inductance in inductors on a PCB when the axial shafts are rotated, which may be translated into angular positions of the joystick.

Abstract: Various embodiments discussed herein enable client applications to be heavily integrated with a voice assistant in order to perform commands associated with voice utterances of users via voice assistant functionality and also seamlessly cause client applications to automatically perform native functions as part of executing the voice utterance. Such heavy integration also allows particular embodiments to support multi-modal input from a user for a single conversational interaction. In this way, client application user interface interactions, such as clicks, touch gestures, or text inputs are executed alternative or in addition to the voice utterances.

Abstract: A payload is generated by a build computing system. The payload includes payload identifying metadata and deployment target identifier metadata, as well as an application container that includes a set of signed binaries. The payload can be approved for deployment, and a token can be generated for the payload. The token may include a set of claims that include payload identifying information. A payload validation system validates the payload for deployment to a target environment using the payload identifying information.

Abstract: This document relates to creating and/or updating a chatbot using a graphical user interface. For example, training dialogs for a chatbot can be displayed in a tree form on a graphical user interface. Based at least on interactions between a developer and the graphical user interface, the training dialogs can be modified in the tree form, and training dialogs can be updated based on the modifications provided on the tree form via the graphical user interface.

Abstract: A computing device is provided, including a plurality of memory devices, a plurality of direct memory access (DMA) controllers, and an on-chip interconnect. The on-chip interconnect may be configured to implement control logic to convey a read request from a primary DMA controller of the plurality of DMA controllers to a source memory device of the plurality of memory devices. The on-chip interconnect may be further configured to implement the control logic to convey a read response from the source memory device to the primary DMA controller and one or more secondary DMA controllers of the plurality of DMA controllers.

Abstract: A system includes a stored counter value and a stored boot manifest including a manifest type flag. A manifest type of the boot manifest is determined based on the manifest type flag, a tenancy mode is determined based on a parity of the counter value, a first boot is executed if the manifest type is a first manifest type and the tenancy mode is a first tenancy mode, a second boot flow is executed if the manifest type is the first manifest type and the tenancy mode is a second tenancy mode, a third boot flow is executed if the manifest type is a second manifest type and the tenancy mode is the first tenancy mode, and a fourth boot flow is executed if the manifest type is the second manifest type and the tenancy mode is the second tenancy mode.

Abstract: Techniques for generating a high resolution full color output image from lower resolution sparse color input images are disclosed. A camera generates images. The camera's sensor has a sparse Bayer pattern. While the camera is generating the images, IMU data for each image is acquired. The IMU data indicates a corresponding pose the camera was in while the camera generated each image. The images and IMU data are fed into a motion model, which performs temporal filtering on the images and uses the IMU data to generate a red-only image, a green-only image, a blue-only image, and a monochrome image. The color images are up-sampled to match the resolution of the monochrome image. A high resolution output color image is generated by combining the up-sampled images and the monochrome image.

Abstract: Techniques for evaluating multiple images, which originate from multiple different sources, and for selecting specific images to generate an overlaid image are disclosed. A first set of system camera images (e.g., based on a first FPS rate) and a second set of external camera images (e.g., based on a second FPS rate) are obtained. A set of rules are accessed in order to govern a selection process for selecting a specific system camera image and a specific external camera image. The selected images are designated for use in generating an overlaid image. The selection process is performed using the accessed set of rules. The overlaid image is generated by overlaying and aligning content obtained from the selected images.

Abstract: Examples described in this disclosure relate to gating a semiconductor layer into a quantum spin Hall insulator state, Certain examples further relate to using quantum spin Hall insulators as topological quantum qubits. Quantum spin Hall systems may rely upon the quantum spin Hall effect by causing a state of a matter to change from a certain phase to an inverted bandgap phase. In one example, the present disclosure relates to a device including a semiconductor layer comprising an active material. The device further includes a gate coupled to the semiconductor layer, where the semiconductor layer is operable in a quantum spin Hall insulator state by using electrons and holes from the active material in response to an application of an electric field to the semiconductor layer via the gate.

Abstract: A code completion tool uses a neural transformer model with attention to generate candidate sequences to complete a method body of a method signature. The neural transformer model is trained with source code programs and natural language text. The neural transformer model learns the meaning of a method name, its corresponding method parameters and types from a large corpus of unsupervised dataset of source code methods and a supervised dataset of tasks including source code constructs in combination with natural language docstrings to infer a candidate sequence of subtokens that represent a method body for a particular method signature.

Abstract: Embodiments detect risky situations in which a domain name record remains viable after the target it identified is not. Such dangling records create various risks because substitute targets, such as fraudulent websites, may be installed without the knowledge of the original target's owner. By obtaining and correlating data from multiple tenants, a cloud service provider detects dangling structures and any attempts to exploit them. Dangling records may specify a custom domain name, for example, or a static IP address that can be misused. In response, the provider's security infrastructure can alert the original target's owner, block the attempted exploit, or otherwise mitigate the risks. Traffic monitoring, control plane API invocations, and domain name server queries may be employed by the security infrastructure to detect resource deletion, resource creation, and resource access attempts that correspond with vulnerable records or suspect activity involving them.

Abstract: The present disclosure relates to systems, methods, and computer readable media for predicting surplus capacity on a set of server nodes and determining a quantity of deferrable virtual machines (VMs) that may be scheduled over an upcoming period of time. This determination of VM quantity may be determined while minimizing risks associated with allocation failures on the set of server nodes. This disclosure described systems that facilitate features and functionality related to improving utilization of surplus resource capacity on a plurality of server nodes by implementing VMs having some flexibility in timing of deployment while also avoiding significant risk caused as a result of over-allocated storage and computing resources. In one or more embodiments, the quantity of deferrable VMs is determined and scheduled in accordance with rules of a scheduling policy.

Abstract: Examples of the present disclosure describe systems and methods for migrating mailbox identities. In aspects, a system may receive one or more requests to perform tasks for migrating identities of a plurality of mailboxes included in a security group created at a source tenant to a plurality of mail users created at a target tenant that correspond to the plurality of mailboxes. Based on the one or more requests, the system may copy the identities to a data store associated with the target tenant, upload mapping data to the data store, and use that mapping data to map the identities copied to the data store to the corresponding plurality of mail users in the target tenant. For each of the identities mapped, the system may copy at least one or more attributes of the respective identity from the data store to the respective corresponding mail user in the target tenant.

Abstract: A training-stage technique trains a language model for use in an ASR system. The technique includes: obtaining a training corpus that includes a sequence of terms; determining that an original term in the training corpus is not present in a dictionary resource; segmenting the original term into two or more sub-terms using a segmentation resource; determining that the segmentation of the original term into the two or more sub-terms is a valid segmentation, based on two or more validity tests; and training the language model based on the terms that have been identified. A computer-implemented inference-stage technique applies the language model to produce ASR output results. The inference-stage technique merges a sub-term with a preceding term if these two terms are separated by no more than a prescribed interval of time.