Abstract: Methods and systems are provided for generating a customized speech recognition neural network system comprised of an adapted automatic speech recognition neural network and an adapted language model neural network. The automatic speech recognition neural network is first trained in a generic domain and then adapted to a target domain. The language model neural network is first trained in a generic domain and then adapted to a target domain. Such a customized speech recognition neural network system can be used to understand input vocal commands.

Abstract: This disclosure describes one or more implementations of a text sequence labeling system that accurately and efficiently utilize a joint-learning self-distillation approach to improve text sequence labeling machine-learning models. For example, in various implementations, the text sequence labeling system trains a text sequence labeling machine-learning teacher model to generate text sequence labels. The text sequence labeling system then creates and trains a text sequence labeling machine-learning student model utilizing the training and the output of the teacher model. Upon the student model achieving improved results over the teacher model, the text sequence labeling system re-initializes the teacher model with the learned model parameters of the student model and repeats the above joint-learning self-distillation framework. The text sequence labeling system then utilizes a trained text sequence labeling model to generate text sequence labels from input documents.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer-readable media that utilize a neural network having a long short-term memory encoder-decoder architecture to progressively modify a digital image in accordance with a natural language request. For example, in one or more embodiments, the disclosed systems utilize a language-to-operation decoding cell of a language-to-operation neural network to sequentially determine one or more image-modification operations to perform to modify a digital image in accordance with a natural language request. In some cases, the decoding cell determines an image-modification operation to perform partly based on the previously used image-modification operations. The disclosed systems further utilize the decoding cell to determine one or more operation parameters for each selected image-modification operation. The disclosed systems utilize the image-modification operation(s) and operation parameter(s) to modify the digital image (e.g.

Abstract: A Quorum network comprising an access controlled multi-tenant network is provided that is configured to enable access control and state isolation in a multi-tenancy Ethereum-based distributed ledger system. The access-controlled network includes one or more authenticating servers (also referred to as access controls) for providing permission control to the nodes in the network. In a standard multi-tenant network, each user of an entity (also referred to as an organization) is limited to only transacting with (also referred to as accessing) their own authorized resources. The access-controlled network utilizes an access controller to provide a singular truth for a set of managed nodes through a trusted entity (such as a Network Authorization Server).

Abstract: One embodiment provides a method comprising determining metadata corresponding to an input image, determining a black level of a display device and a peak luminance value of the display device, and determining a tone mapping function that adaptively compensates for the black level of the display device based on the metadata, the black level of the display device, and the peak luminance value of the display device. The metadata comprises a distribution of luminance values in the input image. The black level of the display device and the peak luminance value of the display device represent a lowest luminance value and a highest luminance value, respectively, achievable on the display device. The method further comprises applying the tone mapping function to the input image to generate a tone-mapped image that adaptively compensates for the black level of the display device and provided to the display device for presentation.

Abstract: A motor-operated drive system for a window covering system including a headrail, a mechanism associated with the headrail to spread and retract the window covering, and a continuous cord loop extending below the headrail for actuating the mechanism to spread and retract the window covering. The drive system includes a motor, a driven wheel that engages and advances the continuous cord loop, and a coupling mechanism for coupling the driven wheel to a rotating output shaft of the motor for rotation of the driven wheel. The drive system includes a channel system for redirecting the continuous cord loop engaged by the driven wheel, or other mechanism for configuring the drive system so that continuous cord loop extends in a substantially vertical orientation. The coupling mechanism includes an engaged configuration in which rotation of the output shaft of the motor causes rotation of the driven wheel, and a disengaged configuration.

Abstract: A cloud-based migration system exposes a source-independent application programming interface for receiving data to be migrated. The data is uploaded and stored as a single entity in a cloud-based storage system. A migration system then accesses the migration package and begins migrating the data to its destination, from the cloud-based storage system.

Abstract: The invention proposed the thermal imaging radar includes of main components: Assembly pedestal, Assembly rotary shaft, and Assembly housing. Electronic circuits, encoders, mechanisms, motor are optimized arranged and scientifically designed the layout space and the weight of the structure. This device is compact for camouflage purposes, easy to assemble or disassemble, and waterproof. The invention's products can be applied in automatic security station, produces 360-degree panoramic imaging of the continuously day and night for surveillance area, detects and tracks moving objects captured by the thermal sensor. Furthermore, The product is also applicable for monitoring the ambient temperature in large areas, localizing high-temperature areas to recognize and warn the possible explosions.

Abstract: A plurality of processing entities in which a plurality of tasks are executed are maintained. Memory access patterns are determined for each of the plurality of tasks by dividing a memory associated with the plurality of processing entities into a plurality of memory regions, and for each of the plurality of tasks, determining how many memory accesses take place in each of the memory regions, by incrementing a counter associated with each memory region in response to a memory access. Each of the plurality of tasks are allocated among the plurality of processing entities, based on the determined memory access patterns for each of the plurality of tasks.

Abstract: This disclosure describes methods, non-transitory computer readable storage media, and systems that provide a platform for on-demand selection of machine-learning models and on-demand learning of parameters for the selected machine-learning models via cloud-based systems. For instance, the disclosed system receives a request indicating a selection of a machine-learning model to perform a machine-learning task (e.g., a natural language task) utilizing a specific dataset (e.g., a user-defined dataset). The disclosed system utilizes a scheduler to monitor available computing devices on cloud-based storage systems for instantiating the selected machine-learning model. Using the indicated dataset at a determined cloud-based computing device, the disclosed system automatically trains the machine-learning model.

Abstract: One embodiment of a hydraulic system for a machine has a first hydraulic circuit including a first pump coupled to a first hydraulic actuator configured to move a first implement of the machine. A second hydraulic circuit includes a second pump coupled to a second hydraulic actuator configured to move a second implement. An electric motor mechanically couples to the first pump and to the second pump. An operator interface receives input from an operator requesting movement of the first and second implements. A controller communicatively coupled to the electric motor and to the operator interface determines, based on the requested movement of the first and second implements respectively, first and second flow allocations respectively for the first and second pumps and determines respective target displacements for the first and second pumps.

Abstract: The present disclosure describes a model for large scale color prediction of objects identified in images. Embodiments of the present disclosure include an object detection network, an attention network, and a color classification network. The object detection network generates object features for an object in an image and may include a convolutional neural network (CNN), region proposal network, or a ResNet. The attention network generates an attention vector for the object based on the object features, wherein the attention network takes a query vector based on the object features, and a plurality of key vector and a plurality of value vectors corresponding to a plurality of colors as input. The color classification network generates a color attribute vector based on the attention vector, wherein the color attribute vector indicates a probability of the object including each of the plurality of colors.

Abstract: A device, computer implemented method, computer program and non-transitory computer-readable storage, for automatically generating negative samples for training a knowledge graph embedding model, The method includes providing at least one first triple, the first triple is a true triple of a knowledge graph, providing at least one second triple, training the knowledge graph embedding model to predict triples of the knowledge graph depending on a set of triples comprising the at least one first triple and the at least one second triple, determining vector representations of entities and relations with the knowledge graph embedding model, determining a plurality of triples with the vector representations of entities and relations, providing an ontology comprising constraints that characterize correct triples, determining with the ontology at least one triple that violates at least one constraint of the constraints or that violates a combination of at least some of the constraints.

Abstract: Implementations relate to ascertaining to what extent predictions, generated using a machine learning model, can be effectively reconstructed from model updates, where the model updates are generated based on those predictions and based on applying a particular loss technique (e.g., a particular cross-entropy loss technique). Some implementations disclosed generate measures that each indicate a degree of conformity between a corresponding reconstruction, generated using a corresponding model update, and a corresponding prediction. In some of those implementations, the measures are utilized in determining whether to utilize the particular loss technique (utilized in generating the model updates) in federated learning of the machine learning model and/or of additional machine learning model(s).

Abstract: A network slice managing entity, comprising means configured to obtain a service request for a network slice, the service request defining service requirements; request network slice descriptions from a database and selecting at least one network slice which support said service requirements according to the descriptions; generate a resource graph comprising nodes representing infrastructure resources and links representing connections between the nodes, the resource graph depicting resources needed for providing the service, wherein the generating is carried out based on a network slice description of the selected at least one network slice; transmit the resource graph to an infrastructure manager for mapping a topology of said network infrastructure onto said resource graph and determining at least one subset of said network infrastructure; and, receive a mapping result from said infrastructure manager containing said at least one subset of said network infrastructure.

Abstract: Systems and methods for natural language processing are described. One or more embodiments of the present disclosure receive a query related to information in a table, compute an operation selector by combining the query with an operation embedding representing a plurality of table operations, compute a column selector by combining the query with a weighted operation embedding, compute a row selector based on the operation selector and the column selector, compute a probability value for a cell in the table based on the row selector and the column selector, where the probability value represents a probability that the cell provides an answer to the query, and transmit contents of the cell based on the probability value.

Abstract: A drive system for raising and lowering a window covering includes a motor, a driven wheel configured to engage a continuous cord loop, a controller for the motor, a sensor, and a housing. The continuous cord loop includes an endless loop of flexible material and one or more sensor targets disposed on the endless loop of flexible material. The housing supports a guide rail adjacent the driven wheel. The sensor is mounted to the guide rail and is configured to generate a signal indicating presence of each sensor target when the target is located in proximity to or in contact with the sensor. The controller is calibrated to store an initial position of each sensor target along the continuous cord loop, and is configured to receive the signal indicating presence of the sensor target and to identify a drift from the initial position during continuing operation of the drive system.

Abstract: A cache memory includes cache lines to store information. The stored information is associated with physical addresses that include first, second, and third distinct portions. The cache lines are indexed by the second portions of respective physical addresses associated with the stored information. The cache memory also includes one or more tables, each of which includes respective table entries that are indexed by the first portions of the respective physical addresses. The respective table entries in each of the one or more tables are to store indications of the second portions of respective physical addresses associated with the stored information.

Abstract: Solid state sources offers potential advantages including high brightness, electricity savings, long lifetime, and higher color rendering capability, when compared to incandescent and fluorescent light sources. To date however, many of these advantages, however, have not been borne out in providing white LED lamps for general lighting applications. The inventors have established that surface recombination through non-radiative processes results in highly inefficient electrical injection. Exploiting in-situ grown shells in combination with dot-in-a-wire LED structures to overcome this limitation through the effective lateral confinement offered by the shell the inventors have demonstrated core-shell dot-in-a-wire LEDs, with significantly improved electrical injection efficiency and output power, providing phosphor-free InGaN/GaN nanowire white LEDs operating with milliwatt output power and color rendering indices of 95-98.

Abstract: Disclosed are compounds of formulae I and II, and pharmaceutically acceptable salts and prodrugs thereof, which are inhibitors of the complement system. Also provided are pharmaceutical compositions comprising such a compound, and methods of using the compounds and compositions in the treatment or prevention of a disease or condition characterized by aberrant complement system activity.