Abstract: A method for instantaneous learning in text-to-speech (TTS) during dialog includes receiving a user pronunciation of a particular word present in a query spoken by a user. The method also includes receiving a TTS pronunciation of the same particular word that is present in a TTS input where the TTS pronunciation of the particular word is different than the user pronunciation of the particular word. The method also includes obtaining user pronunciation-related features and TTS pronunciation related features associated with the particular word. The method also includes generating a pronunciation decision selecting one of the user pronunciation or the TTS pronunciation of the particular word that is associated with a highest confidence. The method also include providing the TTS audio that includes a synthesized speech representation of the response to the query using the user pronunciation or the TTS pronunciation for the particular word.

Abstract: A method for providing response-hiding searchable encryption includes receiving a search query for a keyword from a user device associated with a user. The keyword appears in one or more encrypted documents within a corpus of encrypted documents stored on an untrusted storage device. The method also includes accessing a document oblivious key-value storage (OKVS) to obtain a list of document identifiers associated with the keyword. Each document identifier in the list of document identifiers associated with a respective keyword identifier is concatenated with the keyword and uniquely identifies a respective one of the one or more encrypted documents that the keyword appears in. The method also includes returning the list of document identifiers obtained from the document OKVS to the user device.

Abstract: A method for rehearsal-free continual learning includes obtaining a set of training samples where training sample in the set of training samples is associated with a respective task of a plurality of different tasks. The method includes obtaining a task-invariant prompt representative of learned knowledge common to each respective task of the plurality of different tasks. The method includes, for each respective task of the plurality of different tasks, obtaining a respective task-specific prompt representative of learned knowledge specific to the respective task. The method includes, during each of one or more training iterations, for each respective training sample in the set of training samples, selecting the respective task-specific prompt representative of the respective task of the respective training sample and training a model using the task-invariant prompt and the selected respective task-specific prompt.

Abstract: A method for load balancing application requests across a multi-cluster container orchestration system includes receiving a load-balancing configuration for a multi-cluster service managing access to a set of destination clusters hosting a software application deployed by a user. The multi-cluster service uses the load-balancing configuration to load balance application level traffic across the set of destination clusters. Each destination cluster includes at least one container executing the software application and a respective geographical region. The method also includes receiving an application level request directed toward the software application hosted across the destination clusters. The application level request is received from a client and includes a host name and a geographical location associated with the client. The method also includes routing the request to one of the clusters based on the geographical location of the request and the respective geographical locations of the clusters.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for obtaining, for each of multiple words or sub-words, audio data corresponding to multiple users speaking the word or sub-word; training, for each of the multiple words or sub-words, a pre-computed hotword model for the word or sub-word based on the audio data for the word or sub-word; receiving a candidate hotword from a computing device; identifying one or more pre-computed hotword models that correspond to the candidate hotword; and providing the identified, pre-computed hotword models to the computing device.

Abstract: This document describes systems and techniques for spatially and temporally dynamic illumination for fingerprint authentication. In aspects, a method is disclosed that involves receiving user input at a display of an electronic device, initiating biometric authentication, and instructing a display-driver integrated circuit (DDIC) to implement a local high-brightness mode, which causes a predetermined portion of the display to increase in luminance to a target luminance for at least a first interval and a second interval. The DDIC is also instructed to cause a variable portion of the display to increase in luminance for the first interval and, at the expiration of the first interval, decrease luminance. During the first and/or second interval, an under-display fingerprint sensor captures one or more images. It is then determined, based on analysis of the one or more images, whether the user input is indicative of an authorized user.

Abstract: A method for training a neural network includes receiving a training input audio sequence including a sequence of input frames defining a hotword that initiates a wake-up process on a user device. The method further includes obtaining a first label and a second label for the training input audio sequence. The method includes generating, using a memorized neural network and the training input audio sequence, an output indicating a likelihood the training input audio sequence includes the hotword. The method further includes determining a first loss based on the first label and the output. The method includes determining a second loss based on the second label and the output. The method further includes optimizing the memorized neural network based on the first loss and the second loss associated with the training input audio sequence.

Abstract: Techniques are described herein that enable advanced gaming and virtual reality control using radar. These techniques enable small motions and displacements to be tracked, even in the millimeter or submillimeter scale, for user control actions even when those actions are optically occluded or obscured.

Abstract: The various implementations described herein include methods and systems for determining device leadership among voice interface devices. In one aspect, a method is performed at a first electronic device of a plurality of electronic devices, each having microphones, a speaker, processors, and memory storing programs for execution by the processors. The first device detects a voice input. It determines a device state and a relevance of the voice input. It identifies a subset of electronic devices from the plurality to which the voice input is relevant. In accordance with a determination that the subset includes the first device, the first device determines a first score of a criterion associated with the voice input and receives second scores of the criterion from other devices in the subset. In accordance with a determination that the first score is higher than the second scores, the first device responds to the detected input.

Abstract: This document describes methods and devices for a handover of a user equipment (110) from a Fifth Generation (5G) New Radio (NR) base station (121) to an Evolved Universal Terrestrial Radio Access (E-UTRA) base station (122). The user equipment (110) may determine to release or keep a Packet Data Convergence Protocol (PDCP) entity (312) depending on the whether the E-UTRA base station (122) is connected to an Evolved Packet Core (160) network or a Fifth Generation Core (150) network. The user equipment (110) may determine to release or keep the PDCP entity (312) depending based on a received NR Radio Resource Control (RRC) message or E-UTRA RRC message.

Abstract: A method for forecasting time-series data, when executed by data processing hardware, causes the data processing hardware to perform operations including receiving a time series forecasting query from a user requesting a time series forecast forecasting future data based on a set of current time-series data. The operations include obtaining, from the set of current time-series data, a set of training data. The operations include training, using a first portion of the set of training data, a first sub-model of a forecasting model and training, using a second portion of the set of training data, a second sub-model of the forecasting model. The second portion is different than the first portion. The operations include forecasting, using the forecasting model, the future data based on the set of current time-series data and returning, to the user, the forecasted future data for the time series forecast.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for speech recognition. One of the methods includes receiving first audio data corresponding to an utterance; obtaining a first transcription of the first audio data; receiving data indicating (i) a selection of one or more terms of the first transcription and (ii) one or more of replacement terms; determining that one or more of the replacement terms are classified as a correction of one or more of the selected terms; in response to determining that the one or more of the replacement terms are classified as a correction of the one or more of the selected terms, obtaining a first portion of the first audio data that corresponds to one or more terms of the first transcription; and using the first portion of the first audio data that is associated with the one or more terms of the first transcription to train an acoustic model for recognizing the one or more of the replacement terms.

Abstract: This document describes techniques directed to a scalable gesture sensor for wearable and soft electronic devices. The scalable gesture sensor is integrated into an object such as a wearable garment or a large-surface embedded system to provide a touch-sensitive surface for the object. The sensor includes a repeated localized crossover pattern formed by the same few sensor lines, resulting in the same two conductive lines having multiple crossover points across the touch-sensitive surface. The repeated crossover pattern enables detection of the occurrence and relative direction of a swipe gesture based at least on a repeated sequence of capacitance changes over a set of conductive lines in the repeated crossover pattern. Also, the scalable gesture sensor is computationally simple, uses low power, and is uniquely scalable to cover a large area with few electrodes.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for identifying a user in a multi-user environment. One of the methods includes receiving, by a first user device, an audio signal encoding an utterance, obtaining, by the first user device, a first speaker model for a first user of the first user device, obtaining, by the first user device for a second user of a second user device that is co-located with the first user device, a second speaker model for the second user or a second score that indicates a respective likelihood that the utterance was spoken by the second user, and determining, by the first user device, that the utterance was spoken by the first user using (i) the first speaker model and the second speaker model or (ii) the first speaker model and the second score.

Abstract: A crosstalk cancellation filter set configured for use in delivering binaural signals to human ears is provided. The crosstalk cancellation filter set includes a pressure matching system configured to perform spatial filtering or sound field control and an obstructed field model in communication with the pressure matching system. The crosstalk cancellation filter set is configured to take acoustic advantage of scattering effects and occlusional effects caused by violations to a free-field assumption, thereby delivering improved crosstalk cancellation acoustic displays to a listener without the use of headphones.

Abstract: A method includes receiving a current spectrogram frame and reconstructing a phase of the current spectrogram frame by, for each corresponding committed spectrogram frame in a sequence of M number of committed spectrogram frames preceding the current spectrogram frame, obtaining a value of a committed phase of the corresponding committed spectrogram frame and estimating the phase of the current spectrogram frame based on a magnitude of the current spectrogram frame and the value of the committed phase of each corresponding committed spectrogram frame in the sequence of M number of committed spectrogram frames preceding the current spectrogram frame. The method also includes synthesizing, for the current spectrogram frame, a new time-domain audio waveform frame based on the estimated phase of the current spectrogram frame.

Abstract: In some implementations, a language proficiency of a user of a client device is determined by one or more computers. The one or more computers then determines a text segment for output by a text-to-speech module based on the determined language proficiency of the user. After determining the text segment for output, the one or more computers generates audio data including a synthesized utterance of the text segment. The audio data including the synthesized utterance of the text segment is then provided to the client device for output.

Abstract: This document describes techniques and devices for blood-solute calculation with a mobile device using non-invasive spectroscopy. A mobile device (502) includes a light source (504) that emits light toward an interferometer (508) that uses mirrors to separate and recombine the light. The interferometer directs the recombined light toward a person. Light reflected from, or transmitted through, the person is received through a reception port (506) to a photodetector (510) that outputs photodetector data that corresponds to a measured light intensity of the reflected and transmitted light as a function of a path length of the light or a mirror position of the interferometer. Based on the photodetector data, an interferogram is generated. Applying a technique such as a Fourier transform to the interferogram, a spectrum data set of the reflected and transmitted light is generated. Based on the spectrum data set, a concentration of solutes in the person's blood is calculated.

Abstract: In some implementations, (i) audio data representing a voice command spoken by a speaker and (ii) a speaker identification result indicating that the voice command was spoken by the speaker are obtained. A voice action is selected based at least on a transcription of the audio data. A service provider corresponding to the selected voice action is selected from among a plurality of different service providers. One or more input data types that the selected service provider uses to perform authentication for the selected voice action are identified. A request to perform the selected voice action and (i) one or more values that correspond to the identified one or more input data types are provided to the service provider.

Abstract: A method includes obtaining a speaker identification (SID) model trained to predict speaker embeddings from utterances spoken by different speakers, the SID model includes a trained audio encoder and a trained SID head. The method also includes receiving a plurality of synthetic speech detection (SSD) training utterances that include a set of human-originated speech samples and a set of synthetic speech samples. The method also includes training, using the trained audio encoder, a SSD head on the SSD training utterances to learn to detect the presence of synthetic speech in audio encodings encoded by the trained audio encoder. The operations also include providing, for execution on a computing device, a multitask neural network model for performing both SID tasks and SSD tasks on input audio data in parallel.

Abstract: Techniques and devices for multicast address subscriptions in a wireless mesh network are described in which a mesh network device generates a Multicast Listener Discovery, MLD, Report message including a multicast address. The mesh network device sends the generated MLD Report message to a network interface and a Wireless Personal Area Network, WPAN, Tun interface intercepts the MLD Report message. The mesh network device parses the intercepted MLD Report message to extract the multicast address, translates the intercepted MLD Report message into a mesh network Application Programming Interface, API, function call to join the multicast address, and invokes the API function call on a mesh network interface to transmit a multicast join request message over the mesh network.

Abstract: The present document describes an electronic device with a structural midframe and associated methods. The architectural design of the electronic device (e.g., a security camera) is such that its components are assembled onto the midframe to form a subassembly and the housing is assembled after the subassembly. The midframe includes various features that enable multiple printed circuit boards, a camera subassembly, a front housing member, a heatsink, and a heat spreader to be assembled onto the midframe outside of the housing. The midframe can also include a hinge-bearing surface forming a portion of a ball joint for supporting rotational movement of the electronic device. Accordingly, the electronic device uses the midframe, rather than the housing, as a structural member.

Abstract: This document describes apparatuses and techniques for disposing an underfill dam on a device adjacent to a mounting location for an integrated circuit to direct a flow of underfill toward the mounting location so that the underfill does not flow in undesired directions that may undesirably affect adjacent components.

Abstract: This document discloses aspects of function call authorization for program flow control. In some aspects, a processor encounters a first instruction to initiate or call a function. The processor compares an immediate value of a second instruction at an entry point of the function to a function call authorization value stored in a register. In response to the immediate value of the second instruction matching the function call authorization value stored in the register the process transfers control flow to the function. Alternatively, if the values do not match, an exception or fault may be raised to halt execution of the function or other code. By so doing, these and other aspects of function call authorization may prevent fault injection attacks, execution of unauthorized instructions, or access to sensitive data.

Abstract: A method of encoding and decoding user interface interactions includes receiving a query from a user captured by an assistant-enabled device associated with the user. The query includes a user intent for interacting with an application. The method includes obtaining, for the application, a set of categorized actions. Each categorized action of the set of categorized actions is associated with one or more parameters and represents a high-level description of the user intent of the user interacting with the application. For each respective categorized action of the set of categorized actions, the method includes selecting a respective script associated with the respective categorized action that performs one or more low-level interactions with the application and executing the respective script to represent the user intent for interacting with the application.

Abstract: This document describes techniques and apparatuses directed to force distributing spring contacts. The disclosed spring contacts electrically connect a first component to a second component within a computing device. In implementations, the computing device includes a spring contact having a flexure coupled to a contact pin and configured to apply a force to the second component through the contact pin. The contact pin can include a small or precise structure configured to provide a precise electrical connection to the second component. However, the force to the second component through the small or precise structure of the contact pin increases an applied pressure, which can damage the second component. The disclosed techniques and apparatuses describe force distributing spring contacts with a force distributor configured to distribute the force to the second component over a larger surface area, thereby decreasing the applied pressure and preventing damage to the second component.

Abstract: Techniques and apparatuses are described for determining a position of user equipment by using adaptive phase-changing devices. In aspects, a base station transmits wireless signals for a UE toward respective reconfigurable intelligent surfaces (RISs) of adaptive phase-changing devices. The APDs may direct reflections of the wireless signals in a direction, such as toward the UE, based on a configuration of the RIS of the APD. The base station receives, from the UE via a wireless connection identifiers of the reflections of the wireless signals that are received by the UE. In some cases, the base station also receives a signal quality parameter associated with the reflection reaching the UE. The base station determines angular information based on the respective identifiers and/or signal quality parameters of the reflections. Based on the angular information and known positions of the APDs, the base station determines a position of the UE.

Abstract: A method for detecting an injection vulnerability of a client-side templating system includes receiving a web page, determining that the web page implements an interpreted programming language framework with client-side templating, and extracting a version of the interpreted programming language framework and an interpolation sign from the web page. The method also includes generating an attack payload for at least one injection vulnerability context of the web page based on the version of the interpreted programming language framework and the interpolation sign, instrumenting the web page to inject the attack payload into the at least one injection vulnerability context of the web page, and executing the instrumented web page.

Abstract: Provided herein are systems and methods for performing dynamic adaption and correction for internal delays in devices connected to a common time-multiplexed bus. The methods allow devices to operate reliably at a higher bus frequency by correcting for inherent and unknown delays within the components and in the system by measuring the actual delays using multiple readings with the bus. Intrinsic noise and jitter are used to increase the precision of the measurements, thereby essentially using these uncertainties as self-dithering for increased measurement resolution. During adaption, delays may be adjusted in multiple step sizes to speed adaption time.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for speaker diarization are disclosed. In one aspect, a method includes the actions of receiving audio data corresponding to an utterance. The actions further include determining that the audio data includes an utterance of a predefined hotword spoken by a first speaker. The actions further include identifying a first portion of the audio data that includes speech from the first speaker. The actions further include identifying a second portion of the audio data that includes speech from a second, different speaker. The actions further include transmitting the first portion of the audio data that includes speech from the first speaker and suppressing transmission of the second portion of the audio data that includes speech from the second, different speaker.

Abstract: A method for estimating an embedding capacity includes receiving, at a deterministic reference encoder, a reference audio signal, and determining a reference embedding corresponding to the reference audio signal, the reference embedding having a corresponding embedding dimensionality. The method also includes measuring a first reconstruction loss as a function of the corresponding embedding dimensionality of the reference embedding and obtaining a variational embedding from a variational posterior. The variational embedding has a corresponding embedding dimensionality and a specified capacity. The method also includes measuring a second reconstruction loss as a function of the corresponding embedding dimensionality of the variational embedding and estimating a capacity of the reference embedding by comparing the first measured reconstruction loss for the reference embedding relative to the second measured reconstruction loss for the variational embedding having the specified capacity.

Abstract: Techniques and apparatuses for recognizing accented speech are described. In some embodiments, an accent module recognizes accented speech using an accent library based on device data, uses different speech recognition correction levels based on an application field into which recognized words are set to be provided, or updates an accent library based on corrections made to incorrectly recognized speech.

Abstract: A method for searchable encryption with a public key includes receiving an operation request from a user device associated with a user requesting that encryption of data associated with the user. The data includes a corpus of documents stored on a remote storage device. The method also includes receiving a public key associated with the user. The public key includes an asymmetric cryptographic public key. The method also includes generating a random data key. The data key includes a symmetric cryptographic key. The method also includes encrypting, using the data key, a search index for the corpus of documents based on keywords within the corpus of documents. The method also includes encrypting, using the public key, the data key and sending the encrypted data key to a user device associated with the user.

Abstract: Methods and systems for automatically transitioning content based on facial recognition, wherein the method may include receiving a signal from a camera, identifying a face of a viewer within an area, retrieving a user identifier associated with the face of the viewer, selecting a program, and displaying the program.

Abstract: This document describes techniques and systems that enable an input-mode notification for a multi-input mode. A radar field enables an electronic device to accurately determine the presence or absence of a user near the electronic device. Further, the electronic device can determine whether an application that can receive input through radar-based gestures is operating on the electronic device. Using these techniques, the electronic device can present an input-mode notification on a display of the electronic device when the user is near to alert the user that radar gesture input is available for interacting with the electronic device. This allows the device to provide the user with feedback that can educate the user about what the electronic device is capable of, and allows the user to take advantage of the additional functionality and features provided by the availability of the radar gestures.

Abstract: A method, system and computer program product for workflow style governance includes creating a document and specifying a document style of the document -- either a linear workflow style or a branched workflow style. Thereafter, in response to a subsequent request to load the document, the specified document style for the document may be retrieved and, on condition that the specified document style is a linear workflow style, a specified version of the document is retrieved into a document editor for editing as a new version of the document. But otherwise, on condition that the specified document style is a branched workflow style, different components of the document are retrieved as a composition of components, each of an independent version, and a selected one of the components of the composition of the components making up the document is then loaded into a component editor for editing as a new version.

Abstract: In an aspect, a computer implemented method receives, for a volume within a building, a set of images captured from corresponding locations within the volume, each image taken at a corresponding location, and each image having data specifying the corresponding location from which the image was captured; receiving a three-dimensional point cloud of the volume; detects, using one or more computer vision processing techniques, items of interest within the images; for each item of interest in each image, determines, based on the point cloud and the data specifying the corresponding locations from which the images were captured, a coordinate for the item of interest, the coordinate specifying a location within the volume that corresponds to the item of interest in the image; and stores the data specifying each item of interest and, for each item of interest, its determined coordinate, in a data storage.

Abstract: A computer-implemented method when executed by data processing hardware of a user device causes the data processing hardware to perform operations. Die operations include obtaining, from a message server, an encrypted message encrypted by a single-use data encryption key (DEK) and an encrypted DEK including the single-use DEK encrypted by a public key (PK). The operations also include transmitting, to a key access control list server (KACLS), a decryption request requesting the KACLS decrypt the encrypted DEK with a PRK associated with the PK. The decryption request includes the encrypted DEK. The KACLS is independent from the message server. The operations also include receiving, from the KACLS, the single-use DEK and decrypting, using the single-use DEK, the encrypted message.

Abstract: A distributed storage system including memory hosts and at least one curator in communication with the memory hosts. Each memory host has memory, and the curator manages striping of data across the memory hosts. In response to a memory access request by a client in communication with the memory hosts and the curator, the curator provides the client a file descriptor mapping data stripes and data stripe replications of a file on the memory hosts for remote direct memory access of the file on the memory hosts.

Abstract: A cable assembly includes a connector head electrically coupled to a plurality of conductors disposed in a cable, the connector head having a plug and a conductor interface; an outer shell including a first portion encasing the plug of the connector head, and a second portion encasing the conductor interface of the connector end; an inner shell including a connector end encasing the second portion of the outer shell, and a cable end encasing a portion of the cable and including an injection gate; and conductive resin including a first portion disposed inside the connector end of the inner shell, and a second portion disposed inside and outside the cable end of the inner shell; wherein inside and outside portions of the second portion of the conductive resin are connected through the injection gate.

Abstract: Techniques and/or apparatuses receive an indication that a user has entered a rating of first media content, determine, responsive to the indication that the user has entered the rating of the first media content, whether the user consumed the first media content prior to entering the rating. Responsive to a determination that the user did not consume the first media content prior to entering the rating, the techniques and/or apparatuses provide an indication that the user did not consume the first media content prior to entering the rating or weight the rating based on the determination that the user did not consume the first media content prior to entering the rating of the first media content.

Abstract: The present document describes battery zero-voltage (0V) detection methodologies and applications thereof. These techniques detect the occurrence of the 0V condition and enable the device to continue operation after one 0V condition. These techniques may, however, shut down the device after detecting a second occurrence of the 0V condition of the battery. In an aspect, a combination of hardware and software is used at a system side of a device, when the device is connected to an external power source, to detect a 0V condition and/or distinguish the 0V condition from an undervoltage-protection (UVP) condition.

Abstract: Methods, devices, systems, and means for selection of a coordinating user equipment, UE, in a user equipment-coordination set, UECS, that facilitates selection of the coordinating UE by the UEs in the UECS are described herein. Sharing of the role of coordinating UE by UEs in the UECS can be scheduled in a round-robin manner. Selection criteria can be used to determine which one or more UEs in the UECS will offer better performance in the role of coordinating UE.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-readable storage media, for speech recognition using attention-based sequence-to-sequence models. In some implementations, audio data indicating acoustic characteristics of an utterance is received. A sequence of feature vectors indicative of the acoustic characteristics of the utterance is generated. The sequence of feature vectors is processed using a speech recognition model that has been trained using a loss function that uses a set of speech recognition hypothesis samples, the speech recognition model including an encoder, an attention module, and a decoder. The encoder and decoder each include one or more recurrent neural network layers. A sequence of output vectors representing distributions over a predetermined set of linguistic units is obtained. A transcription for the utterance is obtained based on the sequence of output vectors. Data indicating the transcription of the utterance is provided.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for speech endpointing based on word comparisons are described. In one aspect, a method includes the actions of obtaining a transcription of an utterance. The actions further include determining, as a first value, a quantity of text samples in a collection of text samples that (i) include terms that match the transcription, and (ii) do not include any additional terms. The actions further include determining, as a second value, a quantity of text samples in the collection of text samples that (i) include terms that match the transcription, and (ii) include one or more additional terms. The actions further include classifying the utterance as a likely incomplete utterance or not a likely incomplete utterance based at least on comparing the first value and the second value.

Abstract: This document describes techniques and systems that enable a range extender device. The techniques and systems include a user device that includes a housing with an audio sensor, a heat sink assembly, a circuit board assembly, and a speaker assembly positioned within the housing. The housing includes a top housing member connected to a bottom housing member. The top housing member includes a concave-down top-end portion connected to a generally-cylindrical vertical wall via rounded corners. The heat sink assembly includes a heat sink and one or more antennas positioned proximate to an inner surface of the vertical wall. The circuit board assembly is positioned within the housing and proximate to the heat sink assembly, and the speaker assembly is positioned within the housing and connected to the circuit board assembly. Also, a light ring assembly is connected to a bottom exterior surface of the bottom housing member.

Abstract: A method includes receiving transform coefficients corresponding to a scaled video input signal, the scaled video input signal including a plurality of spatial layers that include a base layer. The method also includes determining a spatial rate factor based on a sample of frames from the scaled video input signal. The spatial rate factor defines a factor for bit rate allocation at each spatial layer of an encoded bit stream formed from the scaled video input signal. The spatial rate factor is represented by a difference between a rate of bits per transform coefficient of the base layer and an average rate of bits per transform coefficient. The method also includes reducing a distortion for the plurality of spatial layers by allocating a bit rate to each spatial layer based on the spatial rate factor and the sample of frames.

Abstract: A method includes receiving, from an image capture device in communication with the data processing hardware, image data for an area of interest of a user. The method also includes receiving a query from the user referring to one or more objects detected within the image data and requesting a digital assistant to discern insights associated with the one or more objects referred to by the query. The method also includes processing the query and the image data to: identify, based on context data extracted from the image data, the one or more objects referred to by the query; and determine the insights associated with the identified one or more objects for the digital assistant to discern. The method also includes generating, for output from a user device associated with the user, content indicating the discerned insights associated with the identified one or more objects.

Abstract: A method of distributing data in a distributed storage system includes receiving a file, dividing the received file into chunks, and determining a distribution of the chunks among storage devices of the distributed storage system based on a maintenance hierarchy of the distributed storage system. The maintenance hierarchy includes maintenance levels, and each maintenance level includes one or more maintenance units. Each maintenance unit has an active state and an inactive state. Moreover, each storage device is associated with a maintenance unit. The determining of the distribution of the chunks includes identifying a random selection of the storage devices matching a number of chunks of the file and being capable of maintaining accessibility of the file when one or more maintenance units are in an inactive state. The method also includes distributing the chunks to storage devices of the distributed storage system according to the determined distribution.

Abstract: A method of sampling output audio samples includes, during a packet loss concealment event, obtaining a sequence of previous output audio samples. At each time step during the event, the method includes generating a probability distribution over possible output audio samples for the time step. Each sample includes a respective probability indicating a likelihood that the corresponding sample represents a portion of an utterance at the time step. The method also includes determining a temperature sampling value based on a function of a number of time steps that precedes the time step, and an initial, a minimum, and a maximum temperature sampling value. The method also includes applying the temperature sampling value to the probability distribution to adjust a probability of selecting possible samples and randomly selecting one of the possible samples based on the adjusted probability. The method also includes generating synthesized speech using the randomly selected sample.