Abstract: A machine learning service implements programmatic interfaces for a variety of operations on several entity types, such as data sources, statistics, feature processing recipes, models, and aliases. A first request to perform an operation on an instance of a particular entity type is received, and a first job corresponding to the requested operation is inserted in a job queue. Prior to the completion of the first job, a second request to perform another operation is received, where the second operation depends on a result of the operation represented by the first job. A second job, indicating a dependency on the first job, is stored in the job queue. The second job is initiated when the first job completes.

Abstract: In various implementations, provided are systems and methods for reducing neural network processing. A compiler may generate instructions from source code for a neural network having a repeatable set of operations. The instructions may include a plurality of blocks. The compiler may add an overwrite instruction to the plurality of blocks that, when executed by one or more execution engines, triggers an overwrite action. The overwrite action causes the instructions of subsequent blocks to be overwritten with NOP instructions. The overwrite action is triggered only when a condition is satisfied.

Abstract: This disclosure describes, in part, techniques for sharing content associated with network applications. For instance, a user may want to share content for a network application, such as a game stream for a gaming application. As such, system(s) may launch a broadcasting session on a first virtual server and launch the network application on a second virtual server. The first virtual server may then receive content data representing states of the network application from the second virtual server. Additionally, the first virtual server may receive video data representing the user and/or audio data representing user speech from a user device. The first virtual server may then generate broadcasting data using the content data, the video data, and the audio data. After generating the broadcasting data, the system(s) may send the broadcasting data to one or more computing devices associated with a user account.

Abstract: A system that performs wall detection and localization to determine a position of a device relative to acoustically reflective surfaces. The device emits an audible sound including a frequency modulated signal and captures reflections of the audible sound. The frequency modulated signal enables the device to determine an amplitude of the reflections at different time-of-arrivals, which corresponds to a direction of the reflection. The device then performs beamforming to generate a 2D intensity map that represents an intensity of the reflections at each spatial location around the device. The device detects wall(s) in proximity to the device by identifying peak intensities represented in the 2D intensity map. In some examples, instead of performing beamforming, the device can perform directional wall detection by physically rotating the device and emitting the audible sound in multiple directions. The device may perform ultrasonic wall detection using ultrasonic sound frequencies.

Abstract: Generally described, systems and methods are provided for monitoring and detecting causes of failures of network paths. The system collects performance information from a plurality of nodes and links in a network, aggregates the collected performance information across paths in the network, processes the aggregated performance information for detecting failures on the paths, analyzes each of the detected failures to determine at least one root cause, and initiates a remedial workflow for the at least one root cause determined. In some aspects, processing the aggregated information may include performing a statistical regression analysis or otherwise solving a set of equations for the performance indications on each of a plurality of paths. In another aspect, the system may also include an interface which makes available for display one or more of the network topology, the collected and aggregated performance information, and indications of the detected failures in the topology.

Abstract: A vacuum-gripper end effector conforms to and grasps items upon application of vacuum force and/or actuation of one or more actuators. The vacuum-gripper assembly can be round in top view or can be rectangular. At least three linear actuators deform the vacuum-gripper assembly. Four linear actuators may be spaced equidistantly apart on a square or rectangular vacuum-gripper assembly to enhance functionality of the vacuum-gripper, such as being capable of simultaneously engaging three sides formed at a corner of a box. A one-way actuation type air cylinder may be used. A low friction air cylinder may be returned to its retracted position by the resilience of the vacuum-gripper assembly.

Abstract: A database system may add a read-only query engine to perform read-only queries associated with points-in-time of a database. In various embodiments, the read-only query engine may be added in response to a manual request, an automatic refresh of a network endpoint, a query specifying a point-in-time, or a connection request. The read-only query engine may perform the point-in-time queries on a version the database at the point-in-time and return results for the queries. Upon completion of the queries or at a determined time, the database system may remove the read-only query engine. The specified point-in-time may refer to a current time, a prior time, or a future time with respect to the current time.

Abstract: Techniques for label recognition and notification for streaming video from non-overlapping cameras. A stream processing service of a provider network receives a first video stream from a first camera-equipped electronic device via an API endpoint of the stream processing service. The stream processing service also receives a second video stream from a second camera-equipped electronic device an API endpoint of the stream processing service. Meanwhile, a request for label recognition and notification is received at a computer vision service of the provider network via an API endpoint of the computer vision service. In response, the computer vision service recognizes a label in a video fragment of the first camera video stream and recognizes a label in a video fragment of the second camera video stream, and then identifies whether the two labels are the same label. If so, a notification service of the provider network sends a notification indicating that the label was recognized across non-overlapping cameras.

Abstract: The systems and methods are provided for the dynamic processing of API requests received by an API execution service. The API execution service may map out a dependency graph based on dependency relationships between all the actions involved in processing the API requests, some of which can be determined from the API definitions and configurations. From the dependency graph, an execution plan can be generated that represents a request processing pipeline conveying the optimal order and arrangement to perform the actions (e.g., serially, in parallel). The execution plan can be followed to process API requests and its performance monitored. The API execution service may dynamically modify this execution plan used to process API requests as it becomes apparent that the execution plan is no longer optimal.

Abstract: Described implementations monitor potential voltage at a location to determine device usage at the location. The implementations utilize a plug-in energy sensor that is plugged directly into any electrical outlet at the location and measures deviation in voltage at the location. Once plugged into an electrical outlet, the plug-in energy sensor monitors one or more of the positive line and ground and/or the neutral line and ground for changes in potential voltage at the location. The plug-in energy sensor may also inject a load (resistive load, inductive load, capacitive load) into the electrical circuit at the location and then measure the signal or response to the injected load.

Abstract: Methods, systems, and computer-readable media for accurate usage forecasting for virtual contact centers are disclosed. A contact center management system configures a contact center instance associated with a client. The contact center instance comprises a plurality of queues configured to store contacts. At least a portion of the contacts are routed to a plurality of agents. The contact center management system determines, using one or more machine learning models associated with the contact center instance, a plurality of predictions for a plurality of metrics for the contact center instance for a plurality of time horizons. At least a portion of the predictions are generated for individual queues of the contact center instance.

Abstract: Devices and techniques are generally described for machine learning hardware optimization. In some examples, a first computing device may receive first data describing a first machine learning model. A first operator type and a first input size for a first layer of the first machine learning model may be determined from the first data. First executable code may be generated that defines a first operator for the first layer of the first machine learning model. The first operator may be specific to the first input size and the first operator type. The first executable code may be stored in non-transitory computer-readable memory. In some examples, second data may be input into the first machine learning model. The first machine learning model may process the second data to generate first output data based at least in part on execution of the first code.

Abstract: Methods and apparatus for client-directed placement of remotely configured service instances are described. One or more placement target options are selected for a client of a network-accessible service based on criteria such as service characteristics of the placement targets. The selected options, including a particular placement target that includes instance hosts configurable from remote control servers, are indicated programmatically to the client. A determination is made that a service instance is to be configured at the particular placement target on behalf of the client. A remote control server is configured to issue administrative commands to an instance host at the particular placement target to configure the service instance.

Abstract: Technologies are disclosed for interacting with a virtual assistant to coordinate, recommend and perform actions. According to some examples, a user may use their voice to interact with a virtual assistant to receive recommendations relating to determining when to perform one or more actions. For example, a user may interact with a virtual assistant to request a recommendation as to when they should leave for the office, leave the office for the day, perform a task, and the like. The recommendation system accesses selected data sources (e.g., calendars, task lists, traffic, transportation schedules, maps, . . . ) to obtain data used in generating the recommendation. In addition, to providing a recommended time, the virtual assistant may also recommend actions to perform. The virtual assistant may also provide notifications to one or more other users that includes information relating to the user leaving.

Abstract: Methods and systems for rendering lists of instructions and performing actions associated with those lists are described herein. In some embodiments, an individual may request that a voice activated electronic device associated with their user account assist in performing a task using a list of instructions. The list of instructions may include metadata that indicates actions capable of being performed by additional Internet of Things (“IoT”) devices. When the instructions are rendered, an instructions speechlet may recognize the metadata and may cause one or more of the IoT devices to perform a particular action. Furthermore, the metadata may also correspond to content capable of being rendered by the voice activated electronic device to assist the individual in performing a particular step of the instructions.

Abstract: Systems and methods are provided to perform multiply-accumulate operations of at least one normalized number in a systolic array. The systolic array can obtain a first input and detect that the first input is denormal. Based on determining the first input is denormal, the systolic array can generate a first normalized number by normalizing the first input. Processing elements of the systolic array can include a multiplier and an adder. The multiplier can multiply the first normalized number by a second normal or normalized number to generate a multiplier product and the adder can add an input partial sum to the multiplier product to generate an addition result.

Abstract: Techniques for input adaptation from disparate data sources for heterogeneous machine learning model execution are described. A preprocessing adapter can perform preprocessing of data obtained from edge devices to suit the input data characteristic requirements of one or more machine learning (ML) models. The preprocessing adapter can determine the input data characteristic requirements in a variety of ways, such as via analysis of the input layer of a ML model or through data variation testing and associated feedback resulting from output data generated by the ML model.

Abstract: Techniques for establishing communication channels between user devices experiencing network connectivity issues and remote communication systems are described herein. The techniques include the use of a secondary device to act as a proxy, or a “middle man,” to facilitate the communications with the user device. A user device may detect lack of network connectivity, and begin broadcasting advertisement messages that indicate the lack of connectivity. A secondary device may detect the advertisement message, and send a discovery message to a connectivity system indicating that it detected the advertisement message. The connectivity system can provide this information to a remote communication system, and the remote communication system can establish a connection with the secondary device and instruct the secondary device to establish a connection with the user device.