Abstract: One or more processors may output for display, an interface including a data classification section including two or more class nodes, a training section including a training node, and an evaluation section including an evaluation node. At a first class node a first set of training data may be captured and at a second class node a second set of training data may be captured. In response to an input received at the training node, a classification model based on the first set of training data and the second set of training data may be trained. Evaluation data may be captured in an evaluation node, and using the trained classification model, classifications for each piece of the evaluation data may be determined. A visual representation of the classification for each piece of the evaluation data may be output for display within the evaluation node.

Abstract: A method includes receiving a training example for a listen-attend-spell (LAS) decoder of a two-pass streaming neural network model and determining whether the training example corresponds to a supervised audio-text pair or an unpaired text sequence. When the training example corresponds to an unpaired text sequence, the method also includes determining a cross entropy loss based on a log probability associated with a context vector of the training example. The method also includes updating the LAS decoder and the context vector based on the determined cross entropy loss.

Abstract: A method for guiding installation of smart-home devices may include capturing, by a camera of a mobile computing device, a view of an installation location for a smart-home device; determining, by the mobile computing device, an instruction for installing the smart-home device at the location; and displaying, by a display of the mobile computing device, the view of the installation location for a smart-home device with the instruction for installing the smart-home device.

Abstract: While texting, a user is able access, share, and control rich media without leaving the texting application. The rich media are provided directly within the executing texting application. The texting application includes an embedded widget for controlling the rich media. Rich media includes, among other things, video clips, streaming audio, a map application, a movie-time application, a social movie-site application, a dynamically controllable image, or promotional media. Different mobile devices executing the texting applications communicate through a server that allows additional functionality, such as syncing the play of video clips and hosting and pushing the promotional media.

Abstract: The various implementations described herein include methods, devices, and systems for implementing high dynamic range and automatic exposure functions in a video system. In one aspect, a method is performed at a video camera device and includes, while operating in a non-high dynamic range (HDR) mode: capturing first video data of a scene with the image sensor; determining whether a minimum number of pixels of the first video data meets one or more first color intensity criteria; and in accordance with the determination that the minimum number of pixels of the first video data meets the one or more first color intensity criteria, switching operation from the non-HDR mode to an HDR mode.

Abstract: A method (500) includes initializing a client state (250) on a client device (120) be executing a private batched sum retrieval instruction (200) to compute c sums O of data blocks (102) from an untrusted storage device (150). Each computed sum O stored on memory hardware (122) of the client device and including a sum of a corresponding subset S of exactly k data blocks. The method also includes a query instruction (300) to retrieve a query block Bq stored on the untrusted storage device by iterating through each of the c sums O of data blocks to identify one of the c sums O that does not include the query block Bq, instructing a service to pseudorandomly partition the untrusted storage device into partitions and sum the data blocks in each partition to determine a corresponding encrypted data block sum (302).

Abstract: A user may routinely wear or hold more than one computing devices. One of the computing devices may be a head-mounted computing-device configured for augmented reality. The head-mounted computing-device may include a camera. While imaging, the camera can consume power and processing resources that diminish a battery of the head-mounted computing device. To improve a battery life and to enhance a user's privacy, imaging of the camera can be deactivated during periods when the user is not interacting with the head-mounted computing device and activated when the user wishes to interact with the head-mounted computing device. The activation of the camera can be triggered by gestured data collected by a computing device other than the head-mounted computing-device.

Abstract: This document describes techniques for fine-motion virtual-reality or augmented-reality control using radar. These techniques enable small motions and displacements to be tracked, even in the millimeter or sub-millimeter scale, for user control actions even when those actions are small, fast, or obscured due to darkness or varying light. Further, these techniques enable fine resolution and real-time control, unlike conventional RF-tracking or optical-tracking techniques.

Abstract: A method for workload security rings that includes receiving a plurality of workloads, each associated with respective security criteria and scheduled for execution on a distributed computing system divided into a plurality of security rings each associated with a respective subset of computing devices of the distributed computing system that is physically isolated from the other security rings. For each respective workload, the method includes determining, using the respective security criteria, a security level of the respective workload and identifying, using the security level of the respective workload, one or more of the plurality of security rings that are eligible for executing the respective workload. The method also includes executing the respective workload on one or more computing devices selected from one of the respective subsets of computing devices associated with the identified one or more of the plurality of security rings eligible for executing the respective workload.

Abstract: Methods, systems, apparatus, including computer programs encoded on computer storage medium, to facilitate language independent-speaker verification. In one aspect, a method includes actions of receiving, by a user device, audio data representing an utterance of a user. Other actions may include providing, to a neural network stored on the user device, input data derived from the audio data and a language identifier. The neural network may be trained using speech data representing speech in different languages or dialects. The method may include additional actions of generating, based on output of the neural network, a speaker representation and determining, based on the speaker representation and a second representation, that the utterance is an utterance of the user. The method may provide the user with access to the user device based on determining that the utterance is an utterance of the user.

Abstract: A method includes detecting, based on sensor data from a sensor on a mobile device, an environmental brightness measurement, where the mobile device comprises a display screen configured to adjust display brightness based on environmental brightness. The method further includes determining, based on image data from a camera on the mobile device, an extent to which the detected environmental brightness measurement is caused by reflected light from the display screen. The method additionally includes setting a rate of exposure change for the camera based on the determined extent to which the detected environmental brightness measurement is caused by reflected light from the display screen.

Abstract: Methods, systems, and media for providing search suggestions are provided. In accordance with some embodiments, the method comprises: receiving user input that includes a partial search query; identifying a group of candidate search suggestions; determining whether each candidate search suggestion is included on a list of search suggestions that are prohibited, wherein the list is generated by: receiving previously submitted search queries and search results; determining, for each of the search queries, a number of search results associated with a particular content rating; determining whether the number meets a threshold value; and in response to determining that the number does not meet the threshold value, inserting the search query to the list; generating a modified group of search suggestions by removing search suggestions that are on the list; and causing a subset of the modified group of search suggestions to be presented as the remaining portion of the search query.

Abstract: A microscope of the type used by a pathologist to view slides containing biological samples such as tissue or blood is provided with the projection of enhancements to the field of view, such as a heatmap, border, or annotations, substantially in real time as the slide is moved to new locations or changes in magnification or focus occur. The enhancements assist the pathologist in characterizing or classifying the sample, such as being positive for the presence of cancer cells or pathogens.

Abstract: Methods, systems and apparatus for implementing a target two-qubit quantum logic gate on a first qubit and second qubit using a tunable qubit coupler. In one aspect, a method includes generating a control signal for the target two-qubit quantum logic gate according to a control model, wherein the control model comprises a controlled-Z operator and a swap operator that are non-orthogonal; and applying the control signal to the first qubit, second qubit and tunable qubit coupler.

Abstract: A Bluetooth audio streaming passthrough, and a method for transmitting an audio stream over a Bluetooth communication link and disclosed. The method is carried out by a source device. An audio codec configuration and a sampling rate of the audio stream are received, the audio stream coming from a streaming application running on the source device. A Bluetooth audio stream between the source device and a sink device is configured using the received audio codec configuration and sampling rate. The first audio stream is received from the streaming application. The received first audio stream is packetized into the Bluetooth audio stream. The Bluetooth audio stream is transmitted to the sink device over the Bluetooth communication link, including the packetized received audio stream.

Abstract: This application discloses an electronic device (e.g., a camera) that a plurality of light sources and a light source driver. The light sources are configurable to a plurality of light source subsets to illuminate a field of view. At least two of the light source subsets include distinct light source members and are configured to illuminate different regions of the field of view. The light source driver is coupled to the plurality of light sources and configured to drive the plurality of light source subsets. In some embodiments, the electronic device includes or is coupled to a camera module configured to capture visual data of the field of view, and the plurality of light sources is configured to provide illumination for the camera module.

Abstract: In an application execution system having a plurality of application servers, each application server stores a plurality of applications, and has computational resources for executing applications in response to received requests. Each application server also includes instructions for loading a respective application into volatile storage and executing the application in response to a request from a client, and for returning a result. A generic application instance may be cloned, creating a pool of generic application instance clones that can be loaded with code for a requested application to produce an application instance. The application instance can then be stored in a cache to be used for a future application request.

Abstract: A method for training a speech recognition model includes obtaining sample utterances of synthesized speech in a target domain, obtaining transcribed utterances of non-synthetic speech in the target domain, and pre-training the speech recognition model on the sample utterances of synthesized speech in the target domain to attain an initial state for warm-start training. After pre-training the speech recognition model, the method also includes warm-start training the speech recognition model on the transcribed utterances of non-synthetic speech in the target domain to teach the speech recognition model to learn to recognize real/human speech in the target domain.

Abstract: A technology for estimating one or more cache hit rates. An implementation includes receiving a request-response pair, calculating a fingerprint for the request-response pair, storing the fingerprint, and determining whether the fingerprint is a member of a bloom filter.

Abstract: A hybrid ANC system that can allow a feedback microphone to receive the same external noise as a feedforward microphone. A processor can generate an anti-noise signal based on what both microphones received to cancel a broader range of frequencies of the external noise.