Abstract: Technologies directed to device detection and locationing are described. One method includes establishing a first wireless connection between a first wireless device and a second wireless device on a first channel. The first wireless device sends to the second wireless device a first indication that the first wireless device is to operate in an off-channel mode with a third wireless device at a first location a Fine Timing Measurement (FTM) procedure. The first wireless device performs a first FTM process between the first wireless device and the third wireless device. The first device determines first timing data including a signal round trip time (RRT) associated with the first FTM process and determines a first distance between the first wireless device and the third wireless device based on the first timing data. A second location of the first wireless device is determined using the first distance.

Abstract: An environmental control system is provided. The environmental control system can include a heat absorption device and a heat rejection device positioned in containment areas and a reservoir. The containment areas can contain leaks that may occur in the environmental control system and direct the fluid into the reservoir. The reservoir can include an opening that can allow the environmental control system to operate as a low pressure system.

Abstract: A system configured to perform distributed echo cancellation processing to attenuate feedback echo from occurring when two devices are acoustically coupled during a communication session. To reduce the feedback echo, one of the devices is configured as a hub device and receives microphone signals, synchronizes the microphone signals, and generates a mixed microphone signal. To enable distributed echo cancellation, the system includes bidirectional feedback link(s) between the hub device and each device synchronized with the hub device. For example, a first bidirectional feedback link sends a microphone signal from a second device to the hub device and sends the mixed microphone signal from the hub device to the second device, which the second device uses to perform echo cancellation. In addition, a second bidirectional feedback link sends a playback signal from the hub device to the second device and sends the output of echo cancellation back to the hub device.

Abstract: Techniques for performing high-precision presence detection by establishing and monitoring peer-to-peer communication links between user devices residing in a same physical environment. A user device that resides in the environment may establish communication links with multiple other user devices residing in the environment. The user device may monitor the channel properties for those communication links to detect fluctuations in the channel properties caused by a user moving through a signal of the communication links. In examples where the user device detects fluctuations in channel properties for multiple communication links, the user device may determine that the user had moved though signals of the communication links. The user device may determine that, because the user moved through signals of multiple communication links, the user may be in close proximity to the user device, such as in the same room.

Abstract: A set of parameters for a set of permissions are determined based at least in part on previous requests to access a set of resources by a principal or user. The set of permissions are updated based at least in part on the set of parameters such that the set of parameters cause different requests to have different authentication requirements. The updated set of permissions is enforced to control access to computing resources such as the set of resources.

Abstract: Techniques are described for generating a specification of security-relevant behavior associated with web services of a cloud provider network. Source code or software development artifacts associated with an implementation of a web service is obtained, where the source code of software development artifacts include an implementation of a request handler for an action of the service. The request handler includes a request authorization component, e.g., which may involve interaction with an identity and access management service of the cloud provider network to authenticate and authorize requests and may further rely upon one or more authorization contexts included in the requests received by the request handler. An interprocedural data flow analyzer is used to analyze a model representation of the bytecode to identify and generate specifications of authorization patterns associated with the request handler.

Abstract: Various techniques for processing sensitive data in an isolated incubator system within a service-provider network are described. The incubator system, for instance, is isolated from a client system in the service-provider network. In an example method, the incubator system receives an indication of an operation, and first encrypted data, from the client system. The incubator system converts the first encrypted data to plaintext and performs the operation. The incubator system converts the processed data into second encrypted data and provides the second encrypted data to the client system. Thus, the incubator system performs the operation on the data without exposing the data to the client system in the plaintext format.

Abstract: From a stream management service, indications of transmissions of data records of a first data stream via a persistent network connection associated with a subscription are provided to a stream processing application. In response to detecting that a transfer throttling condition associated with the subscription has been met, a delay is introduced before additional data records are transmitted via the same persistent network connection. After a subscription expiration period has elapsed, an indication of the expiration is provided to the application via the persistent connection.

Abstract: Embodiments of a system and method for generating a classification model with a cost function having different penalties for false positives and false negatives are described. Embodiments may include perform machine learning operations on known duplicates and known non-duplicates to generate a classification model for classifying structured data items as duplicates or non-duplicates. Each duplicate may represent a pair of structured data items describing a common item; each non-duplicate may represent a pair of structured data items describing different items. Generation of the classification model may be performed based on a cost function that penalizes false positive misclassifications within the classification model differently than false negative misclassifications. Embodiments may also include evaluating the classification model to determine whether a candidate structured data item is a duplicate or non-duplicate.

Abstract: A customer of a resource allocation service can register a function to be executed using virtual resources, where the function includes customer code to be executed. Customer events are defined as triggers for a registered function, and a resource instance is allocated to execute the registered function when triggering event is detected. An identity role associated with the triggering function is used to obtain access credentials for any data source which a triggering event might require for processing. An event-specific access credential is generated that provides a subset of these access privileges using a template policy for the registered function that is filled with values specific to the triggering event. The filled template policy and base credential are used to generate an event-specific credential valid only for access needed for the event. This event-specific credential can be passed with the event data for processing by an allocated instance.

Abstract: In one example, an integrated circuit comprises: a memory configured to store a first mapping between a first opcode and first control information and a second mapping between the first opcode and second control information; a processing engine configured to perform processing operations based on the control information; and a controller configured to: at a first time, provide the first opcode to the memory to, based on the first mapping stored in the memory, fetch the first control information for the processing engine, to enable the processing engine to perform a first processing operation based on the first control information; and at a second time, provide the first opcode to the memory to, based on the second mapping stored in the memory, fetch the second control information for the processing engine, to enable the processing engine to perform a second processing operation based on the second control information.

Abstract: Disclosed are various embodiments for a system that hardware locks a virtual machine to a host machine. The host machine can include a processor and a memory. A hypervisor can be stored in the memory that, when executed by the processor, causes the computing device to host one or more virtual machines. In addition, machine readable instructions stored in the memory that, when executed by the processor, cause the computing device to at least: receive a request to perform an operation on a virtual machine hosted by the hypervisor; evaluate the request to determine that the request complies with a policy; and in response to a determination that the request complies with the policy, cause the hypervisor to initiate the operation.

Abstract: A game-streaming service of a service provider network is configured to assign virtual machine (VM) instances to game streaming requests using a value-based approach. In order to assign a VM instance to a game streaming request for streaming a game to a client device of a player associated with the game streaming request, the game-streaming service may determine one or more attributes of the game streaming request, such as player attributes and/or game attributes. The game-streaming service may also determine available VM instances allocated to a subscriber associated with the game streaming request, which may be located across disparate geographical regions. A value(s) may be determined based on the attributes of the game streaming request, and the value(s) may be used to select one of the available VM instances for streaming the game as part of a corresponding game session.

Abstract: Methods, systems, and computer-readable media for job scheduling based on job execution history are disclosed. A request is received to schedule a workload comprising a plurality of jobs. A resource allocation score for the workload is determined. The resource allocation score represents (at least in part) an estimated likelihood of successful execution of the workload. A first portion of the workload is scheduled for execution, and a remaining portion (if any) of the workload is delayed. A quantity of jobs in the first portion of the workload is determined based (at least in part) on the resource allocation score. Execution of the first portion of the workload is initiated.

Abstract: A matchmaker (e.g., matchmaking software) may be implemented as a number of matchmaker shards, where each shard can potentially run on a different host within a service provider network. Disclosed herein are techniques and systems for dynamically adjusting a shard count associated with a given matchmaker of a subscriber during runtime, the shard count dictating a number of matchmaker shards used for assigning players to matches of a game. Adjustment of the shard count may be based on metrics that are usable to determine whether the current number of matchmaker shards is/are “overloaded”, and, if so, the shard count can be increased, or whether the current number of matchmaker shards are “underloaded”, and, if so, the shard count can be decreased.

Abstract: An application programming interface (API) as a service is disclosed. In embodiments, a client provides code to be executed along with a configuration file for that code. Based on that, virtual machine(s) and load balancer(s) may be selected, a domain name service configured, and throttling and scaling configured. Through this, an API as a service may be provided on behalf of a client with minimal configuration required by the client or an administrator of a web service platform that provides the API as a service.

Abstract: A device may include a display, a housing, a base, and a hinge operably coupling the display and the housing. The hinge may be moveable over a range of positions. One or more loudspeakers are disposed in the housing. A motor couples to the housing and is configured to rotate the housing and the display. The base is configured to remain stationary during actuation of the motor. In some instances, the device may include a camera and one or more microphones for determining a position of a user. The motor may actuate in a direction of the user to provide increase audio and/or visual experiences.

Abstract: Systems and methods for device control by a natural language processing system are disclosed. A user may desire to utilize a voice-enabled device to associate an accessory device with a hub device without having to utilize third-party software associated with the accessory device and/or the hub device. The user may provide a user utterance to associate the accessory device with the hub device. Audio data corresponding to the user utterance may be analyzed and utilized to generate and send directive data to a third-party remote system to transition the hub device to a join mode. Upon association completion, audio may be output confirming that the association has been established successfully.

Abstract: An impact-absorbing package includes an impact absorber and an outer package. An item can be attached to the impact absorber and the impact absorber can be positioned in the outer package. The impact absorber can be positioned to have a volume of air between the item and the bottom of the outer package. An impact force applied to the impact-absorbing package can cause movement between the item and the outer package that is reduced by the impact absorber.

Abstract: A data storage manager may manage storage locations for blocks of a storage volume. The blocks of the storage volume may be assigned to a logical volume exposed to a computing instance supported by a host. Furthermore, the data storage manager may also generate and maintain a set of rules that specify the locations of blocks of the storage volume, and provides the set of rules to the host. The set of rules may be included in a data structure enabling the host to access the blocks based at least in part on the information included in the set of rules.

Abstract: A hub device or edge device implements adaptive data compression. The model training service analyzes the received data. A machine learning model of the hub device receives time-series data from one or more data sources and classifies respective portions of the time-series data as respective patterns. A data compressor at the hub device generates compressed data by applying different compression techniques to the respective portions of the time-series data according to a mapping of the compression techniques to the respective patterns. The hub device then transmits the compressed data to an endpoint for processing (e.g., another device that uses the compressed data). The hub device receives feedback for the compressed data. In response to the feedback, the hub device changes one or more of the compression techniques that are mapped to the respective patterns.

Abstract: A bulk data distribution system in which, instead of multiple data consumers contending to access the same data sets from a primary data store, one or more producers capture snapshots (states of the data sets at particular points in time) and upload the snapshots to an intermediate data store for access by snapshot consumers. The snapshot consumers may download the snapshots to generate local versions of the data sets for access by one or more data processing applications or processes. A snapshot producer may periodically generate full snapshots of a data set, and may generate one or more incremental snapshots of the data set between full snapshots. A snapshot consumer may bootstrap a local data set from a full snapshot and one or more incrementals, and may maintain state of the local data set by accessing new snapshots uploaded by the producer.

Abstract: Systems and methods to grasp objects using vacuum-actuated end of arm tools may include moving pistons, pinching or stabilizing arms, and suction cups. For example, responsive to application of negative pressure and responsive to grasping an object by a suction cup, a piston may move between a retracted position and an extended position. The movement of the piston may cause corresponding movement of one or more pinching or stabilizing arms around the object to pinch and/or stabilize the object grasped by the suction cup.

Abstract: A user may be authenticated at a device through analyzing video frames of acquired video data. Feature recognition may be used to determine that features in the video frames correspond to features of an authorized user. Moreover, successive frames of the video frames may be compared to determine whether the video data is representative of a video capturing the user, or a representation of the user, such as a digital image of or a video depicting the user.

Abstract: Techniques for evaluating a natural language understanding (NLU) component and determining an action to resolve an issue processing a user input are described. The system determines which component is invoked by a baseline NLU component is processing the user input, and which component is invoked by an updated NLU component. Based on that information, the system selects the action to resolve the updated NLU component generating an undesired response to the user input.

Abstract: A design inspector tool generates secure source code related to stencils and design elements of an architecture diagram. The design inspector tool may retrieve source code from a source code repository that includes source code that is relevant to the stencils and design elements implemented by the design inspector tool. When or after a user modifies the source code, the design inspector tool feeds contextual information associated with the stencils and the design elements into a trained machine learning logic. The trained machine learning logic processes the contextual information to retrieve contextually relevant auto complete secure code suggestions from the source code repository. The contextually relevant auto complete source code suggestions may be presented to the user as an option for replacing or augmenting the modified source code.

Abstract: Implementations for providing services to a constrained environment are described. A user interface may be provided for editing files. The files or data structure storing the files may exceed a storage limitation associated with the user interface. The user interface may represent the data structure comprising the files while only storing data indicating modifications to the file or data structure.

Abstract: Disclosed are systems and methods to determine and rank large areas encompassing many parcels (e.g., neighborhoods, cities, towns) for aerial item delivery availability, without the use of image data of the areas. In some implementations, publicly available two-dimensional parcel maps that indicate parcel boundaries and outlines of structures on those parcels may be obtained and processed. For example, parcels within the area may be processed to determine deliverable area shapes, such as rectangles, within the parcel, excluding the area of the structure. A determination is then made as to whether one or more of the deliverable area shapes exceed a deliverable area threshold. If one or more of the deliverable area shapes of the parcel exceed the threshold, the parcel is considered to be available for aerial item delivery. This processing may be done for all parcels within an area or all customer parcels of customers of a service within the area (or any other selection criteria).

Abstract: Disclosed are various embodiments for distributing data items within a plurality of nodes. A data item that is subject to a data item update request is updated from a master node to a plurality of slave notes. The update of the data item is determined to be locality-based durable based at least in part on acknowledgements received from the slave nodes. Upon detection that the master node has failed, a new master candidate is determined via an election among the plurality of slave nodes.

Abstract: A mobile device renders a page that receives input points on a signature provided in conjunction with a mobile payment transaction. The page applies a smoothing algorithm to smooth the input points and create a signature file that more truly represents the actual input signature. The signature file and/or the input points are provided to a payment facilitator for use in completing the mobile payment transaction, performing fraud detection, and potentially for other purposes. Mechanisms are also disclosed for authenticating a buyer device in conjunction with a mobile payment transaction.

Abstract: Methods and apparatus that allow clients to connect resource instances to virtual networks in provider network environments via private IP. Via private IP linking methods and apparatus, a client of a provider network can establish private IP communications between the client's resource instances on the provider network and the client's resource instances provisioned in the client's virtual network via links from the private IP address space of the virtual network to the private IP address space of the provider network. The provider network client resource instances remain part of the client's provider network implementation and may thus also communicate with other resource instances on the provider network and/or with entities on external networks via public IP while communicating with the virtual network resource instances via private IP.

Abstract: A computing device is configured to generate a content reference for content displayed on a webpage. An encrypted reference is generated for the content reference. The content is provided, with the content reference, to a client device. A website is provided for the client device upon selection of the content by a user input on the webpage. The website includes an offering and related offerings associated with the content. In accordance with a web action received from the website, a logical table is updated with an entry associated with the encrypted reference. The update may be associated with a credit or a charge based at least in part on the web action for the offering or the related offerings.

Abstract: A smart container for facilitating operations in a facility is provided. The smart container includes an apparatus configured to store and transport one or more items, one or more electronic devices disposed one or more sides on the apparatus, and a wheel-based charging system coupled to the apparatus and configured to charge one or more battery packs for the one or more electronic devices. The one or more electronic devices are configured to transmit information for monitoring and tracking a position of the smart container throughout an environment.

Abstract: Methods, systems, and computer-readable media for catalog item selection based on visual similarity are disclosed. A similarity detection system selects candidate items for an unavailable item in a catalog. The candidate items are selected from one or more item categories in which the unavailable item is classified. The system determines respective similarity scores for at least some of the candidate items with respect to the unavailable item. For a particular candidate item, the similarity score is determined based at least in part on a visual similarity between an image of the unavailable item and an image of the particular candidate item. The system selects, from the candidate items, a set of visually similar items to the unavailable item based at least in part on the similarity scores. The visually similar items are available to the user.

Abstract: Techniques for performing runtime ranking of skill components are described. A skill developer may generate a rule indicating a skill component is to be invoked at runtime when a natural language input corresponds to a specific context. At runtime, a virtual assistant system may implement a machine learned model to generate an initial ranking of skill components. Thereafter, the virtual assistant system may use skill component-specific rules to adjust the initial ranking, and this second ranking is used to determine which skill component to invoke to respond to the natural language input. Overtime, if a rule results in beneficial user experiences, the virtual assistant system may incorporate the rule into the machine learned model.

Abstract: Described herein is a system for responding to a frustrated user with a response determined based on spoken language understanding (SLU) processing of a user input. The system detects user frustration and responds to a repeated user input by confirming an action to be performed or presenting an alternative action, instead of performing the action responsive to the user input. The system also detects poor audio quality of the captured user input, and responds by requesting the user to repeat the user input. The system processes sentiment data and signal quality data to respond to user inputs.

Abstract: Determining whether to allow access to a message is disclosed. A message is received from a sender. The message is associated with a first time-to-live (TTL) value. A determination is made that the first time-to-live value has not been exceeded. The determination is made at least in part by obtaining an external master clock time. In response to the determination, access is allowed to the message.

Abstract: A computing system providing virtual computing services may maintain a fleet of servers that host virtual machine instances having a wide variety of types and configurations. A service provider may rent processor and memory capacity by defining and offering various virtual machine instances to clients. Each virtual machine instance may include one or more virtual CPUs and a fixed amount of virtualized memory allocated to each virtual CPU, dependent on a predefined ratio between virtual CPU capacity and virtualized memory capacity for the instance type. Each server may include a custom, non-standard sized physical memory module containing memory devices of multiple technologies, types, or sizes on the same printed circuit board. By including custom memory modules, rather than relying only on standard memory modules, the service provider system may implement virtual machines having finer grained options for processor and memory capacity combinations, and may avoid stranding rentable resources.

Abstract: Systems, methods, and computer-readable media are disclosed for ambient light sensing using light guides. In one embodiment, an example device may include a cover layer, a light guide, a light emitting diode disposed adjacent to an edge surface of the light guide, and an ambient light sensor disposed adjacent to the light emitting diode. The ambient light sensor may be configured to sense ambient light that propagates through the cover layer and the light guide.

Abstract: A system comprising one or more computers implements a hardware feature access service. The hardware feature access service stores private keys that correspond to digital certificates embedded in chipsets of devices enrolled in the hardware feature access service. The hardware feature access service is configured to issue access or access revocation messages to the chipsets to “lock” or “unlock” associated hardware components. The hardware feature access service also implements a service interface that allows clients to request changes to enabled feature sets for devices enrolled in the hardware feature access service. In response to such requests, the hardware feature service automatically and wirelessly enables or disables feature sets by locking or unlocking relevant hardware components of a device relevant to enabling or disabling the requested feature sets.

Abstract: Techniques for database performance degradation detection and prevention are described. A statement performance monitor observes queries executed against a database engine and clusters the queries into groups of queries. The index utilization of the query groups and execution metrics are tracked over time, and a sudden change of index utilization can be detected. The change can be reported to users and/or new indexes may be automatically generated to serve affected query groups. Additionally, a statement performance monitor may be deployed to statically analyze code to identify modified queries and the resultant change of use of query indexes.

Abstract: To improve the reliability of a memory system, data and error correction codes associated with the data can be stored in a first memory. Parity bits calculated over data bits in the first memory can be stored in a second memory. The parity bits in the second memory can be used to recover errors that are uncorrectable by the error correction codes. The first memory can be implemented, for example, using an emerging memory technology, while the second memory can be implement using a different memory technology.

Abstract: A global queue service of a provider network allows clients to automatically replicate a queue to another queue of one or more other regions (creating a global queue). A local instance of the service receives a send request from a producer. In response, the local instance stores a message in a local queue and replicates the send request to a remote queue of a remote instance of the global queue service (e.g., at another region). The local instance receives a receive request from a consumer and sends the message to the consumer (without replicating the request to the remote queue). The local instance then receives a delete request from a consumer. In response, the local instance deletes the message from the local queue and replicates the delete request to the remote queue. The local instance also receives replicated requests from the remote instance and in response, applies them locally.

Abstract: In one example, a neural network accelerator executes instructions to: load a first weight data element of an array of weight data elements from a memory into a systolic array; extract, from the instructions, information indicating a first number of input data elements to be obtained from a first address of the memory and a second number of input data elements to be skipped between adjacent input data elements to be obtained, the first address being based on first coordinates of the first weight data element, and the first and second numbers being based on a stride of a convolution operation; based on the information, obtain first input data elements from the first address of the memory; and control the systolic array to perform first computations based on the first weight data element and the first input data elements to generate first output data elements of an output data array.

Abstract: Methods, systems, and computer-readable media for analysis of code coverage differences across environments are disclosed. A code coverage profiling system determines a first code coverage profile associated with execution of program code in a first environment. The first code coverage profile indicates one or more portions of the program code that were executed in the first environment. The code coverage profiling system determines a second code coverage profile associated with execution of the program code in a second environment. The second code coverage profile indicates one or more portions of the program code that were executed in the second environment. The code coverage profiling system performs a comparison of the first code coverage profile and the second code coverage profile. The comparison determines a difference between the portions of the program code that were executed in the first and second environments.

Abstract: A database system may include a database service and a separate distributed storage service. The database service (or a database engine head node thereof) may be responsible for query parsing, optimization, and execution, transactionality, and consistency, while the storage service may be responsible for generating data pages from redo log records and for durability of those data pages. For example, in response to a write request directed to a particular data page, the database engine head node may generate a redo log record and send it, but not the data page, to a storage service node. The storage service node may store the redo log record and return a write acknowledgement to the database service prior to applying the redo log record. The server node may apply the redo log record and other redo log records to a previously stored version of the data page to create a current version.