Abstract: Techniques are described for managing communications between multiple computing nodes, such as for computing nodes that are part of managed virtual computer networks provided on behalf of users or other entities. In some situations, one or more of the computing nodes of a managed virtual computer network is configured to perform actions to extend capabilities of the managed virtual computer network to other computing nodes that are not part of the managed virtual computer network, such as by forwarding communications between computing nodes of the managed virtual computer network and the other external computing nodes so as to enable the other external computing nodes to participate in the managed virtual computer network. In some situations, the computing nodes may include virtual machine nodes hosted on one or more physical computing machines or systems, such as by or on behalf of one or more users.

Abstract: Metadata indicating that a virtual traffic hub enabling connectivity between a plurality of isolated networks has been established is stored. A determination is made that a first entry of a first isolated network attached to the hub is to be represented in a second routing table of a second isolated network attached to the hub, e.g., to enable network packets originating at resources of the second isolated network to be transmitted via the hub to the first isolated network. A new entry corresponding to the first entry is included in the second routing table.

Abstract: Respective correlation metrics between token groups of a particular text attribute of a data set and a prediction target attribute are computed. Based on the correlation metrics, a predictive token group list is created. For various observation records of the data set, values of a derived categorical attribute corresponding to the particular text attribute are determined based on matches between the particular text attribute value and the predictive token group list. A measure of the predictive utility of the particular text attribute is obtained using correlations between the categorical attribute and the prediction target attribute.

Abstract: Methods, systems, and computer-readable media for continuous functions in a time-series database are disclosed. A plurality of data points of a time series are stored into one or more storage tiers of a time-series database. The plurality of data points comprise a plurality of discrete measurements at respective timestamps. Using one or more query processors of the time-series database, a query of the time series is initiated. The query indicates a time range. Using the one or more query processors, a continuous function is determined that represents a segment of the time series in the time range. The continuous function is determined based at least in part on the plurality of data points. An operation is performed using the continuous function as an input.

Abstract: A system comprising one or more computing devices implements a vehicle information extraction service. The vehicle information extraction service enables customers to provide access to proprietary encoding protocols used by the customers to communicate over a bus of a vehicle, wherein the proprietary protocols are used by the vehicle information extraction service to perform vehicle information monitoring, while at the same time protecting the confidentiality of the proprietary protocols from being accessed by third parties. Also, the vehicle information extraction service enables streaming of relevant vehicle information corresponding to customer defined event criteria without requiring non-relevant vehicle information to be streamed, which would consume considerable network resources.

Abstract: A file-level snapshot access service provides direct access to individual files included in a snapshot for virtual volume of a block-storage service without requiring a volume to be re-created from the snapshot, attached to a computing device, or mounted in a file system. For example, a user/client may directly retrieve individual files from specified snapshots via a user interface/API of the file-level snapshot access service. Additionally, the file-level snapshot access service is configured to provide a listing of files included in a given snapshot. In some embodiments, a file-level snapshot access service may provide direct access to individual files included in snapshots generated for other types of storage systems, such as an object-based storage system.

Abstract: Techniques for presence detection are described. In an example, system receives, from a first device connected with a computer network that is associated with a location, a first value of a first parameter associated with a link between the first device and a second device. The system determines, using a first prediction model and the first value, a first likelihood of presence at the location. The first prediction model is configured based at least in part on outputs of a second prediction model that uses a second parameter associated with the computer network. The system determines whether the presence is detected at the location based at least in part on the first likelihood.

Abstract: Disclosed are systems and methods to deliver media files at a highest perceived presented quality. Rather than measure quality as a difference between an encoded version of the media file and the original version of the media file, in accordance with the disclosed implementations, quality is based on a presented quality of each fragment of a media file when presented by a device to a user. The disclosed implementations determine a presented quality score for each fragment of a media file, the presented quality score indicative of a predicted quality of the fragment as perceived by a user when the fragment is presented. The presented quality score may be determined based on one or more of a bitrate used to encode the media segment, a device type of a device used to present the media segment, the content of the segment itself, etc.

Abstract: Techniques for continuous lock-minimal checkpointing and recovery with a distributed log-based datastore are described. A continuous, fault-tolerant checkpoint process is run on a node in a cluster in a compute-optimized or memory-optimized manner, thereby checkpointing the in-memory replica state of the node to a durable checkpoint datastore. A node can partially restore its replica state by obtaining checkpoint data, which includes an identifier of a low-water mark in a journal shard. The node can attach to the journal shard using the low-mater mark as the point of attachment, enabling the node to finalize the restoration of the replica state to be current.

Abstract: An inventory management system can include a fleet of mobile drive units. The mobile drive units can detect, using a sensor, calibration data associated with a calibration fixture. The calibration of the sensors can be updated based on the calibration data. The calibration data for the fleet of mobile drive units can be used determine a correction factor. The correction factor can be applied to the fleet of mobile drive units to further update the calibration of the sensors.

Abstract: Disclosed techniques generate a data structure including data representing a configuration of a data storage system. A parameter can be obtained that is usable to configure the data storage system. Data representing the configuration of the data storage system can be modified based on the parameter. The data storage system can be configured, using the parameter, based on the modified data representing the configuration of the data storage system and predetermined configuration data for the data storage system.

Abstract: A configuration management system tracks the configuration of a managed computing environment in accordance with a configuration tracking policy. The configuration management system uses the tracking policy to map configuration information from the managed computing environment to a hierarchy of configuration items. Items may be included or excluded from the hierarchy based on relevance to the tracking policy and a predicted amount of dynamism. A signature of the hierarchy is generated. A change to the configuration environment is detected based on a change to the signature, and an action is performed in response to the detected change.

Abstract: One or more multi-stage optimization iterations are performed with respect to a compression algorithm. A given iteration comprises a first stage in which hyper-parameters of a perceptual quality algorithm are tuned independently of the compression algorithm. A second stage of the iteration comprises tuning hyper-parameters of the compression algorithm using a set of perceptual quality scores generated by the tuned perceptual quality algorithm. The final stage of the iteration comprises performing a compression quality evaluation test on the tuned compression algorithm.

Abstract: A node disconnection rotation may be performed on an active synchronizing node pool of a multi-node replication system. The node disconnection rotation may ensure that at least one node is disconnected from the active synchronizing node pool at any given time. A disconnected node does not synchronize with other nodes during the time that it is disconnected. A disconnected node may, therefore, serve as a designated survivor, from which the system may be restored, when other active nodes are subject to a cyber-attack. Although the disconnected designated survivor node may potentially be compromised, it may be sanitized of any malicious contents prior to restart.

Abstract: A speech-processing system may provide access to one or more virtual assistants via a voice-controlled device. The system may be activated by detecting a wakeword in speech received by a microphone of the device. The system may process the speech and provide a response in the form of synthetic speech. When a speaker of the device synthetic emits the speech, the microphone may detect some or all of the speech. If the synthetic speech includes a wakeword or words or phrases similar to the wakeword, a wakeword detection component of the device may detect the wakeword and activate an assistant, resulting in a self-wake or cross-wake. Self- or cross-wake may interrupt an action or response currently in progress, which may frustrate the user and result in a poor user experience. This disclosure thus proposes systems and methods for preventing cross-wake and self-wake in a voice-controlled device.

Abstract: This disclosure describes a system for automatically transitioning items from a materials handling facility without delaying a user as they exit the materials handling facility. For example, while a user is located in a materials handling facility, the user may pick one or more items. The items are identified and automatically associated with the user at or near the time of the item pick. When the users enters and/or passes through a transition area, the picked items are automatically transitioned to the user without affirmative input from or delay to the user.

Abstract: Techniques are described for generating transaction data for a transaction, the transaction data generated with varying fidelity levels based on various constraints. In response to a transaction request from a client, a contract engine may perform service call(s) that instruct network service(s) to generate transaction data. The service call(s) may include constraint(s) on the generation of the transaction data, such as a time limit or a requested fidelity. The network service(s) may generate the transaction data at a particular fidelity based on the time limit, requested fidelity, or other criteria such as current load on the network service(s). The transaction data may be communicated to the contract engine with a fidelity indicator. Lower fidelity transaction data may be sent to the client to enable a low latency user experience. The lower fidelity transaction data may subsequently be replaced with higher fidelity transaction data generated with higher latency.

Abstract: A machine learning model, such as a neural network, is partially encrypted with a homomorphic encryption scheme. Application of the machine learning model to data includes performing operations on plaintext and encrypting results of operations for input to other operations that are performed over cyphertext. Ciphertext output of such operations can be provided to a service that is able to decrypt the ciphertext output.

Abstract: Described are systems and methods for detecting objects using calibrated imaging devices and obfuscating, in real-time or near real time, portions of the video data to protect the privacy of operators represented in the video data. For example, a position of an operator within a fulfillment center may be determined or tracked in video data and the pixels representative of that operator may be obfuscated using pixilation and/or other techniques so that a reviewing agent that is viewing the video data cannot determine the identity of the operator. Such obfuscation may be performed in real-time or near real-time using automated processing. In addition, only portions of the video data may be obfuscated so that events (e.g., item picks, item place) and/or other objects represented in the video data are still viewable to the reviewing agent.