Abstract: Embodiments of the present disclosure relate to methods for fabricating optical devices. One embodiment of the method includes disposing a structure material layer on a surface of a substrate and disposing a patterned photoresist over the structure material layer. The patterned photoresist has at least one device portion and at least one auxiliary portion. Each device portion and each auxiliary portion exposes unmasked portions of the structure material layer. The unmasked portions of structure material layer corresponding to each device portion and each auxiliary portion are etched. The etching the unmasked portions forms at least one optical device having device structures corresponding to the unmasked portions of at least one device portion and at least one auxiliary region having auxiliary structures corresponding to the unmasked portions of at least one auxiliary portion.

Abstract: A wireless transceiver. The transceiver includes: (i) a transmit signal path; (ii) a calibration path, comprising a conductor to connect a calibration tone into the transmit signal path; (iii) a receive signal path, comprising a first data signal path to process a first data and a second data signal path, different than the first data signal path, to process a second data; (iv) a first capacitive coupling to couple a response to the calibration tone from the transmit signal path to the first data signal path; and (v) a second capacitive coupling to couple a response to the calibration tone from the transmit signal path to the second data signal path.

Abstract: Methods, systems and computer program products for extensible content object workflows. A content management system stores a plurality of content objects. The content management system integrates applications, which integrated applications make modifications to the content objects based on computer-aided invocation of workflows. Upon receipt of a message from a first integrated application indicating an intent to invoke a workflow at a second integrated application, the content management system relays information pertaining to the desired workflow to the second integrated application. The workflow of the second integrated application modifies one or more of the content objects or its metadata or other data associated with the workflow. Selection of the second integrated application can be based on attributes of the content objects, and/or can be based on user interactions within a user interface.

Abstract: The present disclosure generally relates to systems and methods for implementing power a power split between a first and a second power source in a fuel cell powertrain system. The method includes receiving an input into a processor of the fuel cell powertrain system, determining an output by the processor, communicating the output by the processor to a system controller and determining a power split by the system controller. The first power source includes a fuel cell system and the second power source is selected from a battery system or an engine, and the input includes a life or health of at least one of the first power source or the second power source.

Abstract: Techniques discussed herein facilitate configuration of network slices for a User Equipment (UE) supporting constraints on simultaneous operation of slices. One example embodiment is a UE device configured to: transmit a registration request message to request registration on a set of requested network slices; and receive a registration accept message indicating the UE is registered to a set of allowed network slices, and wherein the registration accept message indicates slice compatibility information for each allowed network slice of the set of allowed network slices for the UE, wherein the slice compatibility information for each allowed network slice indicates whether or not other configured network slices of the set of configured network slices are compatible with the configured network slice for simultaneous operation by the UE.

Abstract: A method of performing speech recognition using a two-pass deliberation architecture includes receiving a first-pass hypothesis and an encoded acoustic frame and encoding the first-pass hypothesis at a hypothesis encoder. The first-pass hypothesis is generated by a recurrent neural network (RNN) decoder model for the encoded acoustic frame. The method also includes generating, using a first attention mechanism attending to the encoded acoustic frame, a first context vector, and generating, using a second attention mechanism attending to the encoded first-pass hypothesis, a second context vector.

Abstract: In one embodiment, an apparatus includes a heat sink for attachment to an optical module cage configured for receiving an optical module, a thermal interface material attached to a surface of the heat sink for thermal contact with the optical module, and a plurality of lifting elements extending from the surface of the heat sink. The lifting elements are configured to create a gap between the thermal interface material and the optical module during insertion of the optical module into the optical module cage or removal of the optical module from the optical module cage, the plurality of lifting elements positioned for insertion into aligned recesses in the optical module when the optical module is fully inserted into the optical module cage to eliminate the gap and provide contact between the optical module and the thermal interface material.

Abstract: In one embodiment, an apparatus includes a fan tray comprising a plurality of fans for cooling modules within a modular electronic system with airflow from a front of the modular electronic system to a back of the modular electronic system, and a hinge member for connecting the fan tray to a front panel of the modular electronic system with the fans positioned in front of the modules. The fan tray is rotatable on the hinge member for movement away from the front panel to allow for replacement of one of the modules.

Abstract: A speech interface device is configured to receive response data from a remote speech processing system for responding to user speech. This response data may be enhanced with information such as remote NLU data. The response data from the remote speech processing system may be compared to local NLU data to improve a speech processing model on the device. Thus, the device may perform supervised on-device learning based on the remote NLU data. The device may determine differences between the updated speech processing model and an original speech processing model received from the remote system and may send data indicating these differences to the remote system. The remote system may aggregate data received from a plurality of devices and may generate an improved speech processing model.

Abstract: Methods of improving adhesion between a photoresist and conductive or insulating structures. The method comprises forming a slot through at least a portion of alternating conductive structures and insulating structures on a substrate. Portions of the conductive structures or of the insulating structures are removed to form recesses in the conductive structures or in the insulating structures. A photoresist is formed over the alternating conductive structures and insulating structures and within the slot. Methods of improving adhesion between a photoresist and a spin-on dielectric material are also disclosed, as well as methods of forming a staircase structure.

Abstract: Systems, methods, and computer-readable media for requesting a cellular IP address by initiating a call with a modem, establishing data packet network connectivity with the cellular IP address, assigning the cellular IP address to a virtual L2-bridge interface, wherein the virtual L2-bridge interface includes a MAC address, mapping a MAC address of a virtual machine with the MAC address of the virtual L2-bridge interface, detecting a change in the cellular IP address, and updating the virtual L2-bridge interface with a different cellular IP address while maintaining the data packet network connectivity.

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

Abstract: Exemplary embodiments relate to techniques for allowing a first user of a messaging platform to send follow-up message(s) to a second user after expiration of an allotted messaging time period without violating existing messaging policies based at least in part on opt-in messaging. The first user may perform an API call to an API of a messaging server to send a follow-up message request. Based on the request, the messaging server may send an opt-in message to the second user who may then explicitly opt in to receive a follow-up message from the first user. A qualifying response by the second user to the opt-in message may cause the messaging server to generate a unique token for the communicating pair. The first user may send the follow-up message to the second user by sending the message with the token.

Abstract: Systems and methods are provided for automatically marking locations within a radiograph of one or more dental pathologies, anatomies, anomalies or other conditions determined by automated image analysis of the radiograph by a number of different machine learning models. Image annotation data may be generated based at least in part on obtained results associated with output of the multiple machine learning models, where the image annotation data indicates at least one location in the radiograph and an associated dental pathology, restoration, anatomy or anomaly detected at the at least one location by at least one of the machine learning models. A number of different pathologies may be identified and their locations marked within a single radiograph image.

Abstract: A photodialysis device that can desalinate saline streams is described herein. The device can have a desalination rate that is significantly faster than current solar thermal desalination technology. Salt is removed from water by passing ionic current derived from sunlight through this water using dye-sensitized membranes. The device can serve as a distributed, low-cost technology that can efficiently and effectively desalinate low salinity sources of water, which is particularly useful for agriculture and drinking water applications.

Abstract: A cellular network includes a first network component configured to i) identify a first session between a first data network and a user equipment (UE), wherein the first session corresponds to a first session management function (SMF), ii) receive data network access information (DNAI) from a network function, the DNAI corresponding to a second data network and iii) select a second SMF that is to be utilized for a second session between the second data network and the UE. The cellular network also includes a second network component configured to i) store a mapping between the DNAI and the second SMF and ii) transmit an indication of the second SMF to the first network component, wherein the first network component selects the second SMF based on the indication.

Abstract: Described are a system, method, and computer program product for reconfiguring a transaction during network processing. The method includes receiving, from a merchant system, a transaction request including transaction data and configured in an initial configuration associated with a single-message transaction or a dual-message transaction. The method includes determining, based at least partly on the transaction data and/or at least one parameter of a merchant profile associated with the merchant system, a recommended configuration associated with a single-message transaction or a dual-message transaction, the recommended configuration different from the initial configuration. The method includes reconfiguring the transaction request based on the recommended configuration.

Abstract: Techniques for validating large amounts of sensitive data migrated across multiple platforms without revealing the content of the sensitive data are provided. For example, a processing device can transform data in a first data file stored on a first platform to common data formats. The processing device can generate a first set of hash values. The processing device can receive a second set of hash values for a second data file stored on a second platform. The processing device can compare the first set of hash values and the second set of hash values and cause the first data file or the second data file to be modified based on a difference between the sets of hash values.

Abstract: An optimal package structure and an amount of refrigerant to insert into the package structure are determined based upon product information, consumer preference information, and weather information. The optimal package structure and the amount of refrigerant are an optimized combination that is effective to both protect the product from damage and preserve the product from spoilage. A package is constructed to contain the product according to the optimal package structure and the amount of refrigerant is inserted into the structure.

Abstract: Disclosed are various embodiments for generating a list of predicted contacts that can be provided to a client device. The predicted contacts can be generated based upon an analysis of user interaction data. The predicted contacts can be made available to a phone application or messaging application on a client device so that contact information can be displayed in response to an incoming call or a message.