Abstract: A speech-processing system receives input data representing text. An encoder processes segments of the text to determine embedding data, and a decoder processes the embedding data to determine one or more categories associated with each segment. Output data is determined by selecting words based on the segments and categories.

Abstract: Systems, methods, and computer-readable media are disclosed for passive throttling of bulk material on conveyor systems. In one embodiment, an example system may include a first conveyor having a first width, a second conveyor having a second width, the second conveyor disposed adjacent to the first conveyor, a third conveyor disposed adjacent to the second conveyor, a chute, and a fourth conveyor. Items falling from the second conveyor may be guided to the fourth conveyor via the chute. The first conveyor, the second conveyor, and the third conveyor may move in a first direction, and the fourth conveyor may move in a second direction.

Abstract: Systems and methods for device-state quality analysis include storing data indicating device state of a device and receiving additional data indicating the device state of the device. Data is then generated based on the stored and received device state data that indicate that the stored data is inaccurate. Confidence value data is then generated utilizing the generated data and prior device-state accuracy data.

Abstract: Systems and methods are disclosed for container dumpers having controlled unloading. In one embodiment, an example container unloading system may include a container support mechanism configured to secure a container, where the container has items inside, and a first conveyor configured to rotate from a first position to a second position. The first conveyor may be disposed in a vertical orientation adjacent to a sidewall of the container in the first position, and the first conveyor may be disposed in a non-vertical orientation adjacent to the sidewall of the container in the second position. The system may include a controller configured to cause the container support mechanism to rotate the container from a vertical orientation to a rotated orientation, and cause the first conveyor to be actuated, such that the items are conveyed out of the container by the first conveyor.

Abstract: This disclosure proposes systems and methods for entity resolution using speech recognition data. The system can receive audio data representing an utterance and perform automatic speech recognition (ASR) processing on the audio data to generate at least a first ASR data and a second ASR data. The system can perform natural language understanding (NLU) processing on the first ASR data to determine intent data and an indication of an entity. The system can determine a first portion of the first ASR data that corresponds to the indication of the entity. The system can determine a second portion of the second ASR data that corresponds to the indication of the entity without performing NLU on the second ASR data. The system can perform entity resolution (ER) on the second portion to identify a first entity.

Abstract: Techniques to improve the operation of homomorphic circuits. Systems and methods described herein may improve the operation of a homomorphic circuit by analyzing a circuit to reduce the multiplicative depth of the circuit. Techniques described herein may be utilized to perform linear algebra operations with mixed encoding units, according to a homomorphic encryption scheme.

Abstract: Techniques for generating synthetic data are described. An exemplary approach includes receiving one or more requests to generate synthetic data based on a first dataset; generating the synthetic dataset is generated according to the request by choosing a set of synthetic datapoints between pairs of datapoints of the first dataset along a line connecting them while sampling a likely value of a local probability distribution; and providing the synthetic dataset as configured by the request.

Abstract: High-fidelity measurements of qubits are achieved by increasing a number of measurements taken by use of a swap operation and a readout qubit, deflating a bosonic qubit for which measurement outcomes are affected by single photon/phonon loss events, deflating a bosonic qubit enabling readout in other basis, and evolving the qubit under a Hamiltonian that couples a mode to be measured to another mode where the Hamiltonian is selected from a three wave mixing interaction, and/or a combination of these techniques.

Abstract: Devices, systems, and methods are provided for using dynamic shipping options. A method may include receiving a request to add a first product to a virtual shopping cart of a first user; determining a first delivery fee for a first order comprising the first product; generating user interface data indicative of the first delivery fee and with which the first user may input a first offered delivery fee; presenting the user interface data; receiving a first user input indicative of the first offered delivery fee offered by the first user; determining that a coincident delivery fee discount based on the first order and a second order is less than a delivery fee threshold including delivery fee discounts of the first offered delivery fee and of a second offered delivery fee of the second order; initiating a first purchase of the first product a second purchase of the second product.

Abstract: Systems and methods of dynamically controlling robots are described. A first robot may be controlled to move a pod to a cart-filling station. A first subset of destinations for the cart-filling station may be determined. First control instructions may be sent to a robotic sortation device at the cart-filling station. The first control instructions may be effective to allocate the first set of destinations to the robotic sortation device. A second robot may be controlled to move a first cart associated with a first destination of the first set of destinations to the cart-filling station. The robotic sortation device may select a first item from the pod associated with the first destination. The robotic sortation device may place the first item in the first cart and the second robot may be controlled to remove the first cart from the cart-filling station.

Abstract: Techniques for feedback-based training may include selecting a scoring machine learning model based at least in part on a test metric, and applying the model on an unlabeled dataset to generate, per dataset item of the unlabeled dataset, a prediction and an importance ranking score for the prediction. Techniques for feedback-based training may further include selecting, based on the importance ranking scores, a result of the application of the model on the unlabeled dataset, providing the result and requesting feedback on the result via a graphical user interface, receiving the feedback via the graphical user interface, adding data from the unlabeled dataset into a training dataset when the feedback indicates a verified result, and retraining the model using the training dataset with the data added from the unlabeled dataset to generate a retrained model.

Abstract: This disclosure is directed to techniques in which a first user in an environment scans visual indicia associated with an item, such as a barcode, before handing the item to a second user. One or more computing devices may receive an indication of the scan, retrieve image data of the interaction from a camera within the environment, identify the user that received the item, and update a virtual cart associated with the second user to indicate addition of the item.

Abstract: A computer-implemented method is disclosed. The method includes selecting one or more target surfaces portrayed in at least one video frame, generating a video data latent space representation of the at least one video frame, accessing a plurality of supplemental data latent space representations of a plurality of supplemental data sets, identifying a particular supplemental data latent space representation based at least in part on the video data latent space representation, selecting a particular supplemental data set in response to identifying the particular supplemental data latent space representation, the particular supplemental data set corresponding with the particular supplemental data latent space representation, and inserting the particular supplemental data set into the at least one video frame.

Abstract: Systems and methods to automatically generate training data for machine learning models may include an imaging device to capture imaging data, an image processing or rendering system to receive the imaging data and render a three-dimensional model of an object of interest overlaying the imaging data, an automatic mask extraction or generation system to extract or determine a mask, label, or annotation associated with the three-dimensional model and a plurality of pixels associated with the object of interest from a perspective of the imaging device, and a machine learning model to receive the imaging data and the mask as training data.

Abstract: Systems and methods for determining whether a first electronic device detects a media item that is to be output by a second electronic device is described herein. In some embodiments, an individual may request, using a first electronic device, that a media item be played on a second electronic device. The backend system may send first audio data representing a first response to the first electronic device, along with instructions to delay outputting the first response, as well as to continue sending audio data of additional audio captured thereby. The backend system may also send second audio data representing a second response to the second electronic device along with the media item. Text data may be generated representing the captured audio, which may then be compared with text data representing the second response to determine whether or not they match.

Abstract: A data deletion notification service registers data repositories that store personal data. The service receives requests from users to delete personal data, sends notifications of the deletion requests to the registered data repositories, and stores deletion records recording the deletions and a timestamp associated with the deletion request. The service may wait for confirmation of the deletions from the data repositories, resending the deletion requests if a confirmation is not received, and recording confirmation in an audit record. Data repositories performing a restoration may request, from the service, a list of users that have requested deletion of personal data since the time of the creation of the backup used for restoration. The service may generate, based on deletion records with corresponding timestamps, the list of users and respond to the request with the list. Confirmations of the deletions may be tracked, stored and made available via an audit interface.

Abstract: An autonomous mobile device (AMD) that interacts with a user operates more effectively when able to persist an entity identifier across time with a particular user. Sensor data, such as a set of image frames, is acquired over time and used to determine locations of the user in the physical space. An entity identifier may be associated with the user as depicted in each frame, to allow the AMD to distinguish one user from another. However, due to changes in orientation, appearance, lighting, and so forth, different entity identifiers may be assigned across different frames, even though the user is the same. Information about the location at different times is used to establish a relationship between a root entity identifier and the entity identifiers in individual frames, allowing for consistent operation with respect to the particular user associated with the root entity identifier.

Abstract: Backup and restore of client-managed tables and system-managed tables in a same database may be performed. A request to restore a database with both a client-managed table and system-managed table may be received. The request may specify a point in time to restore the database. A metadata volume for the database that describes both the client-managed table and the system-managed table may be restored and used to identify the volumes for both the client-managed table and shards of the system-managed table. Redo logs for each of the volumes may be used according to timestamps that are synchronized according to a time synchronization service to generate restored versions of the volumes for the client-managed table and the shards of the system-managed table from backup copies, which can then be made available for accessing the database.

Abstract: Devices and techniques are generally described for unused identity and access management rights detection. In various examples, a first skill-usage vector associated with a first profile may be determined. A first nearest neighbor algorithm and the first skill-usage vector may be used to determine a second skill-usage vector grouped together with the first skill-usage vector in a feature space, where the second skill-usage vector is associated with a second profile. A first rights vector associated with the first profile may be determined. The first rights vector associated with the first profile may be compared to a second rights vector associated with the second profile. At least one unused right associated with the first profile may be determined based at least in part on the comparing of the first rights vector to the second rights vector.

Abstract: Techniques for a location service to determine geographic locations (or “geolocations”) of user devices connected to a telecommunications network, and provide signed certificates that attest to the user devices being at geolocations at particular times. The location service may calculate, or receive, geolocations of user devices that have been determined using network-based localization techniques for telecommunications networks where the network infrastructure is used to determine the geolocations. Because user devices cannot fake or spoof these geolocations, the location service may populate certificates with indications of geolocations of user devices at different times. Additionally, the location service may act as a certificate authority by digitally signing these certificates and issuing digitally signed certificates that attest to geolocations of user devices.

Abstract: Users may enter and exit a facility as individuals, or as groups in which an authorized user sponsors other users. Within the group, actions within the facility taken by those sponsored users may be associated with the account of the authorized user. Sensor data acquired at a portal area is used to determine information about users passing through a portal, such as a crossing time, a distance from a crossing point, and so forth. This information is used to determine group associations. These group associations are resilient to situations, such as employees who have exited and re-entered with a user, employees holding open a gate for a user, people entering ahead of the authorized user, people entering after the authorized user, people exiting before an authorized user, and so forth.

Abstract: Systems and methods to provide environment control to individual containers of a storage system may include a distribution network including one or more sources, distribution lines, valves, sensors, and storage positions. The distribution network may be operated to provide air flow having desired environment attributes to individual storage positions that include supply and/or return lines and associated valves. Insulated containers having input and/or output lines and associated valves may couple to the distribution network upon placement at the storage positions, thereby providing air flow with desired environment attributes to individual insulated containers and associated items. Further, non-insulated containers may also be placed at the storage positions of the storage system without connecting to the distribution network.

Abstract: Systems and methods for performing garbage collection. A system updates a second data structure based at least in part on a first data structure, the first data structure comprising data from a previous garbage collection cycle of a garbage collection application and the second data structure comprising data from a current garbage collection cycle of the garbage collection application. The system allows an application to update the second data structure during the current garbage collection cycle.

Abstract: A system for utilizing media content reference point information to perform media content encoding, and supplemental content stitching and/or insertion. Media content can be encoded and packaged based on boundaries of the media content. The boundaries can be received from a third-party and/or generated via an automated process. Target boundaries can be selected based on accuracy levels associated with the received and/or generated boundaries. Supplemental content can be stitched and/or inserted into packaged media content based on audio and video content of the packaged media content being aligned.

Abstract: A method and system for tracking and determining completion of sub-tasks to ensure completion of an overall task related to an item are described. The method and system receive item information describing a particular item to be acted upon and a task to complete with the item, such as assembly or shipping of the item. One or more intermediate states of the item are determined based on a process for transitioning to the final completed state. One or more intermediate tasks are determined to accomplish the transition to the completed state. The instructions for completing the tasks are then generated and conveyed for completion.

Abstract: Computer systems and associated methods are disclosed to implement the non-interactive join of privacy-preserving dataset sketches. In some embodiments, an entity can publish a one-time sketch of their dataset that would enable another entity to join their data without exposing private information. The sketch can map, using a hash function, the identities associated with a first value of the dataset to a data structure, in some embodiments. A same or different entity can join the first sketch with a privacy-preserving second sketch of a second dataset that includes added noise, and can determine an estimate of a number of identities that correspond with specific values of the first and second datasets from the joined dataset. The sketch can be published just one time, and therefore does not require separate new private computations with privacy budgeting for each additional party when a join is desired, in some embodiments.

Abstract: A geographically distributed data store including a number of geographically distributed regions may be implemented using replication groups that include multiple regions configured according to replication criteria. First tier replication of particular changes to data stored in the distributed data store may be performed in compliance with the replication criteria, where management of replication state is performed with respect to replication across the replication groups. Independent of the first tier replication, individual replication groups may implement second tier replication of changes to data where management of replication state is performed with respect to replication within the particular replication group. Replication group configuration may be determined using the replication criteria which may include thresholds for replication resource utilization, replication latency and utilization of data change logs.

Abstract: A system for applying metadata tags based on metadata tag groups is described. The system includes a plurality of compute nodes configured to provide a plurality of computing resources for clients of a provider network and a resource tag manager for the provider network. The resource tag manager is configured to receive, from a first client of an account of the provider network, a creation request to establish a resource attribute group comprising a plurality of resource keys and a plurality of resource values, wherein respective metadata keys correspond to respective resource values. The resource tag manager may receive, from a second client of the account, an application request to apply the resource attribute group to a computing resource and apply the plurality of resource keys and the plurality of resource values to metadata maintained for the computing resource.

Abstract: Described are systems and methods for determining a reputation score and/or a confidence score for a network identifier that represents, respectively, a likelihood that the network identifier presents a threat and/or a likelihood that the network activity associated with the network identifier corresponds to a port scanning, enumeration, or other malicious event. Embodiments of the present disclosure can utilize various network telemetry information, such as authentication activity, outbound traffic activity, web activity, honeypot connection activity, or network classification information to determine the reputation and/or confidence scores in response to a query/request and/or in connection with potentially malicious activity that can represent a likelihood that the detected potentially malicious activity is malicious/legitimate and the potential impact that remediation measures taken against the network identifier may have in the event that the detected network activity was legitimate.

Abstract: Devices and techniques are generally described for confidence score generation for label generation. In some examples, first data may be received from a first computing device. In various further examples, first label data classifying at least one aspect of the first data may be received. First metadata associated with how the first label data was generated may be received. In some cases, the first label data may be generated by a first user. In various examples, a first machine learning model may generate a first confidence score associated with the first label data based at least in part on the first data and second data related to label generation by the first person. In various examples, output data comprising the first confidence score may be sent to the first computing device.

Abstract: Described are systems and methods for providing a centralized inverter architecture that incorporates multiple inverters co-located in a single integrated power core. Exemplary embodiments of the present disclosure can provide an exemplary power electronics system that includes integrated power core units having multiple inverters co-located in a centralized location. The integrated power cores can enable compute topologies that enable communication between the inverters and coordinated operation of the inverters. The integrated power cores can also be configured as a single line replaceable unit (LRU) to facilitate efficient repair, maintenance, and/or replacement of each integrated power core.

Abstract: Aspects described herein include a computer-implemented method of generating a floor plan for use by a plurality of mobile drive units within an inventory system. The method comprises defining a grid according to predefined grid rules, placing inventory station blocks, placing predefined travel blocks, and placing predefined storage blocks. Predefined connection rules specify which faces of the different predefined blocks are permitted to connect. The method further comprises outputting a map representing the floor plan. The map comprises location information for fiducial markers corresponding to the predefined blocks.

Abstract: A connection management device may be used to with a common API to allow a host device of a client to securely connect to a remote provider network. The host device may only be able to use a connection management device by using a defined set of commands of a common API. A hardware root of trust may be pre-provisioned with security data (e.g., client certificate, encryption keys). A connection command may be used that is not specific to any particular communication protocol (e.g., WiFi, cellular, wired protocol). In response to receiving the connection command from a host device, the connection management device may perform commands specific to the communication protocol of the connection management device to connect to a remote provider network, use the security data for authentication, and establish a connection in accordance with the communication protocol based on the authentication.

Abstract: Methods, systems, and computer-readable media for customizable event-triggered computation at edge locations are disclosed. A request for content is received at an edge server from a client device. The content is sought from a content cache at the edge server or from an origin server coupled to the edge server. Processing of the request is initiated, comprising encountering an event. The event is associated with a function specified by a customer. The function associated with the event is executed at the edge server using process isolation. The content is generated based at least in part on execution of the function. The content is sent from the edge server to the client device.

Abstract: Systems and techniques for detecting vulnerabilities and exploits within network assets in a near real-time manner. The systems provide for scanning and probing assets within an organization network based on determined or disclosed exploits from the wild. The system provides for digital risk protection that actively and passively hunts for potential organizational threats and vulnerabilities and generates alerts in real-time related to the detected exploits. The systems provide a platform for generating plugins to scan for the detected exploits and fingerprint assets for services hosted and potential vulnerabilities.

Abstract: A system configured to perform sound source localization (SSL) using reflection detection is provided. A device processes audio data from multiple microphones to determine timing information corresponding to sound sources. For example, the device may determine cross-correlation data for each microphone pair, determine autocorrelation data for each microphone, and then use the autocorrelation data and the cross-correlation data to calculate quality factors. The device may determine the direction of potential sound source(s) by generating Steered Response Power (SRP) data using the cross-correlation data. To perform reflection detection to distinguish between direct sounds and acoustic reflections, the device may generate modified SRP data using the quality factors.

Abstract: Devices and techniques are described for augmented reality furniture layout recommendation. In some examples, first data comprising a three dimensional (3D) model of a room is received. Second data representing an intended usage of the room may be received. A first machine learning model may generate a first collection template using the first data and the second data. The first collection template may include first bounding box data representing a first category of furniture and second bounding box data representing a second category of furniture. The first bounding box data representing a first bounding box having a first size and located at a first position within the 3D model at a first orientation. The second bounding box data representing a second bounding box having a second size and located at a second position within the 3D model at a second orientation.

Abstract: A computer system may provide a plurality of load records for presentation in a user interface provided at a first plurality of user devices. The user interface may enable searching for a plurality of loads represented by the plurality of load records. The computer system may also determine, for a load record of the plurality of load records and based at least in part on inputting contextual information into a predictive model, an acceptance prediction indicating a probability that a load record of the plurality of load records will be accepted by a user. The computer system may also provide at least a portion of the plurality of load records for presentation in the user interface provided at a second plurality of user devices associated with a second plurality of users. The portion of the plurality of load records may include the load record including an updated value of an attribute of the load record that is dependent on the acceptance prediction.

Abstract: Described are systems, methods, and apparatus that utilize multiple sensors to conserve battery life of a device, limit potential disruption to a user, and reliably detect whether the device is worn by the user. In one example, a first sensor, such as an infrared sensor, may be activated to determine a proximity of the device to an object. If the device is within a proximity of the object, a second sensor, such as a photoplethysmography (PPG) sensor, may be activated to verify that the device is being worn by the user and then to continue collecting data that may be used to determine biometrics for the user.

Abstract: A separate relationship management interface for applications that utilize application data objects to implement various application features is used to track changes to relationships between application data objects. Requests that cause writes corresponding to relationships between application data objects and requests that cause reads of relationships between application data objects may be handled using an interface that interacts with a separate data store tracking the relationships between application data objects.

Abstract: Systems and techniques are disclosed for testing the functional equivalence of a legacy system, such as mainframe system, and an adapted system, such as a cloud version of the mainframe system. The same input can be provided to both the legacy system and the adapted system, such that the legacy system and the adapted system can both operate separately on copies of the same input. Results produced by the legacy system and the adapted system based on the same input can accordingly be compared. If results produced by the legacy system and the adapted system over time are determined to be functionally equivalent, the adapted system can be determined to be functionally equivalent to the legacy system.

Abstract: A data stream conversion service may crawl a non-relational database to determined database attributes that characterize the data structure of the non-relational database. The data stream conversion service may evaluate the one or more attributes of the non-relational database to generate a schema for a relational database, wherein the schema maps the non-relational database attributes to respective fields of a relational database. The data stream conversion service may generate a relational database using the determined schema and replicate a data stream for the non-relational database to the relational database.

Abstract: Systems and methods for presence-based application invocation are disclosed. For example, presence of a personal device and/or user associated with a personal device is detected at a primary device. When account data associated with the personal device and the primary device correspond, checks are performed to determine if automatic invocation of an application associated with the personal device should be performed. Those checks include determining whether one or more conditions are satisfied, determining current usage of the application and/or the primary device indicates desirability to automatically invoke the application, and/or whether the application has been enabled for use on the primary device.

Abstract: Systems and methods for improving training processes for image and text applications are described. A first set of embeddings may be generated based on a text input, and a second set of embeddings may be generated via a convolutional neural network (CNN), based on an input image. The first set of embeddings and the second set of embeddings may be utilized to generate a third set of embeddings including one or more placeholder values to be replaced. The placeholder values may be replaced based on predicted values, to reconstruct the input text and image.

Abstract: Described are systems and methods directed to a body change journey that utilizes one or more current body measurement values, one or more goal body measurement values, and a cohort to determine a body change journey that a user can follow to progress from the current body measurement value to the goal body measurement value. The cohort may include a plurality of data of other bodies who have progressed from a starting body measurement value that is similar to the current body measurement value.

Abstract: Methods, systems, and computer-readable media for compliance lifecycle management for cloud-based resources are disclosed. A selection is received of a compliance pack from a plurality of compliance packs. The compliance pack comprises a plurality of rules associated with policy compliance. The compliance pack is selected from the plurality of compliance packs via a user interface. The selection is associated with one or more resources hosted in one or more provider networks. An evaluation is performed of compliance of the one or more resources with respect to the plurality of rules of the compliance pack. Data describing the evaluation is generated and displayed. The data comprises an aggregate compliance status for at least one of the one or more resources, and the aggregate compliance status represents an aggregate compliance with the plurality of rules.

Abstract: An electronic book distribution system includes electronic devices that reset their passcodes after specified authentication failures. The passcodes of an individual electronic device is reset to a value that is generated using a predefined function of a randomly generated support code. The support code is displayed to the user, and the user is instructed to contact a support service in order to obtain the new passcode. The support service independently authenticates the user, calculates the new device passcode using the same predefined function used by the electronic device, and provides the new passcode to the user.