Abstract: A client-server system that performs machine learning based information fusion to predict part failure likelihood is described. The system receives transactional data pertaining to replacement of, and sensor data pertaining to duty cycle of, one or more parts. The system trains a first machine learning model, using the transactional data as training data, to extract a plurality of concepts corresponding to the information present in unstructured text in the transactional data. The system also trains a second machine learning model, using the sensor data and the extracted plurality of concepts, to predict part failure likelihood of the one or more parts. The system determines the part failure likelihood of the one or more parts by providing new transactional data and new sensor data to the trained machine learning models.

Abstract: The disclosed concepts relate to implementation of application and application engine functionality using machine learning. One example method involves obtaining a seed image representing a seeded application state and mapping the seed image to at least one seed image token using an image encoder. The example method also involves inputting the at least one seed image token as a prompt to a neural dreaming model that has been trained to predict training sequences obtained from one or more executions of one or more applications, the training sequences including images output by the one more applications during the one or more executions and inputs to the one or more applications during the one or more executions. The example method also involves generating subsequent image tokens with the neural dreaming model, and decoding the subsequent image tokens with an image decoder to obtain subsequent images.

Abstract: Systems, devices, methods, and software for sanitization monitoring of hands, other body parts, and objects and training users on sanitization methods. The systems and devices including a detector to provide images of the object within its detection range, and at least one processor to receive the images from the detector, determine areas of the image corresponding to sanitized areas of the object from unsanitized areas of the object, calculate a percentage of sanitized areas to the total area corresponding to the sanitized and unsanitized areas, and report at least the percentage of sanitized area and/or compare a user's sanitization techniques to approved sanitization techniques and provide guidance to user. In various embodiments, users sanitize their hands with fluorescing hand sanitizer and/or a fluorescing germ-proxy agent with soap and water and the sanitized and unsanitized areas are determined based on the amount of fluorescing material remaining on the hands after application.

Abstract: Aspects of the present disclosure provide systems and methods which utilizes machine learning techniques to provide enhanced accessibility features to a game. An accessibility service is provided which is capable of instantiating one or more machine learning models which can process current gameplay states and generate commands to assist users during gameplay. The accessibility commands may be provided to a game and used to supplement or modify user provided inputs in order to compensate for specific user needs. In further aspects, an accessibility user interface is provided which allows a user to dynamically enable or disable accessibility features during gameplay. The user interface is operable to receive accessibility selections and provide the selection data to an accessibility service during gameplay.

Abstract: An in vitro tissue plate may include a well plate, a fluidic plate disposed on a bottom surface of the well plate, and a media manifold disposed on a bottom surface of the fluidic plate. The well plate may have at least two wells, including a tissue well and a waste well. The fluid plate may include a fluid channel extending between and fluidly connecting the tissue well to the waste well. The media manifold may include a one or more media outlets fluidly connected to the fluid channel. A tissue layer may be deposited in the tissue well. The tissue layer may include human cells such as neurovascular cells.

Abstract: A client-server system that performs machine learning based information fusion to predict part failure likelihood is described. The system receives transactional data pertaining to replacement of, and sensor data pertaining to duty cycle of, one or more parts. The system trains a first machine learning model, using the transactional data as training data, to extract a plurality of concepts corresponding to the information present in unstructured text in the transactional data. The system also trains a second machine learning model, using the sensor data and the extracted plurality of concepts, to predict part failure likelihood of the one or more parts. The system determines the part failure likelihood of the one or more parts by providing new transactional data and new sensor data to the trained machine learning models.

Abstract: An apparatus has a memory storing a reinforcement learning policy with an optimization component and a data collection component. The apparatus has a regularization component which applies regularization selectively between the optimization component of the reinforcement learning policy and the data collection component of the reinforcement learning policy. A processor carries out a reinforcement learning process by: triggering execution of an agent according to the policy and with respect to a first task; observing values of variables comprising: an observation space of the agent, an action of the agent; and updating the policy using reinforcement learning according to the observed values and taking into account the regularization.

Abstract: An apparatus has a memory storing a reinforcement learning policy with an optimization component and a data collection component. The apparatus has a regularization component which applies regularization selectively between the optimization component of the reinforcement learning policy and the data collection component of the reinforcement learning policy. A processor carries out a reinforcement learning process by: triggering execution of an agent according to the policy and with respect to a first task; observing values of variables comprising: an observation space of the agent, an action of the agent; and updating the policy using reinforcement learning according to the observed values and taking into account the regularization.

Abstract: A scalable automated maintenance optimisation system for managing replacement of parts in a portfolio of assets is described. A plurality of maintenance records pertaining to one or more replaced parts is received. The received plurality of maintenance records is analysed to generate an asset history including information on replacement of parts. An age at replacement distribution is generated for each part type based on the asset history. A part replacement strategy for each part type is determined based at least on the calculated age at replacement distribution. A user interface is controlled to display a list of the parts and the respective determined part replacement strategy. The displayed list automatically reconfigured by moving parts whose determined part replacement strategy is different from a current part replacement strategy to a position closer to the top of the displayed list of parts in the user interface.

Abstract: In various examples there is a method performed by a computer-implemented agent in an environment. The method comprises storing a reinforcement learning policy for controlling the computer-implemented agent. The method also comprises storing a distribution as a latent representation of a belief of the computer-implemented agent about at least one other agent in the environment. The method involves executing the computer-implemented agent according to the policy conditioned on parameters characterizing the distribution.

Abstract: Systems, devices, methods, and software of the present invention provide for sanitization monitoring of hands, other body parts, and objects. The systems and devices including a detector to provide images of the object within its detection range, and at least one processor to receive the images from the detector, determine areas of the image corresponding to sanitized areas of the object from unsanitized areas of the object, calculate a percentage of sanitized areas to the total area corresponding to the sanitized and unsanitized areas, and report at least the percentage of sanitized area. In various embodiments, users sanitize their hands with fluorescing hand sanitizer and/or a fluorescing germ-proxy agent with soap and water and the sanitized and unsanitized areas are determined based on the amount of fluorescing material remaining on the hands after application.

Abstract: Examples are disclosed that relate to computing device input systems. In one example, a computing device input system comprises an input configured to receive a connection to a control device connector comprising a plurality of conductors. The computing device input system further comprises circuitry configured to determine a presence, type, and state of a control device in communication with the control device connector based on analog voltages received from the control device connector.

Abstract: An apparatus is described for training a behavior of an agent in a physical or digital environment. The apparatus comprises a memory storing the location of at least one reward token in the environment. The location has been specified by a user. At least one processor executes the agent in the environment according to a behavior policy. The processor is configured to observe values of variables comprising: an observation of the agent, an action of the agent and any reward resulting from the reward token. The processor is configured to update the behavior policy using reinforcement learning according to the observed values.

Abstract: An apparatus has a memory storing a reinforcement learning policy with an optimization component and a data collection component. The apparatus has a regularization component which applies regularization selectively between the optimization component of the reinforcement learning policy and the data collection component of the reinforcement learning policy. A processor carries out a reinforcement learning process by: triggering execution of an agent according to the policy and with respect to a first task; observing values of variables comprising: an observation space of the agent, an action of the agent; and updating the policy using reinforcement learning according to the observed values and taking into account the regularization.

Abstract: An apparatus is described for training a behavior of an agent in a physical or digital environment. The apparatus comprises a memory storing the location of at least one reward token in the environment. The location has been specified by a user. At least one processor executes the agent in the environment according to a behavior policy. The processor is configured to observe values of variables comprising: an observation of the agent, an action of the agent and any reward resulting from the reward token. The processor is configured to update the behavior policy using reinforcement learning according to the observed values.

Abstract: Systems for managing content in a cloud-based service platform. A server in a cloud-based environment is interfaced with storage devices that hold one or more stored objects accessible by two or more users. The stored objects comprise folders and files as well as other objects such as workflow objects that are associated with the folders or the files. The workflow objects comprise workflow metadata that describes a workflow as a set of workflow tasks to be carried out in a progression. Processing of a workflow task and/or carrying out a portion of the progression includes modification of shared content objects. The processing or modification events are detected through workflow events, which in turn cause one or more workflow responses to be generated. Workflow responses comprise updates to the workflow metadata to record progression through the workflow and/or workflow responses comprise updates to any one or more of the stored objects.

Abstract: A dual-orientation stand for supporting a user input device in two device orientations is provided. The stand may comprise a first planar surface configured to rest on a support surface in a first stand orientation, and a second planar surface extending from the first planar surface at an obtuse angle. A third planar surface is spaced from the second planar surface and extends from the first planar surface. A concave portion between the second and third surfaces is configured to hold the device in a first device orientation when the stand is in the first stand orientation. The second planar surface is configured to support the device in a second device orientation when the stand is in a second stand orientation in which the third planar surface rests on the support surface.

Abstract: A system for classifying thyroid nodule tissue as malignant or benign is provided that is based on the identification of sets of gene transcripts, which are characterized in that changes in expression of each gene transcript within a set of gene transcripts can be correlated to with either malignant or benign thyroid nodule disease. The thyroid classification system provides for sets of “thyroid classifying” target sequences and further provides for combinations of polynucleotide probes and primers derived there from. These combinations of polynucleotide probes can be provided in solution or as an array. The combination of probes and the arrays can be used for diagnosis. The invention further provides further methods of classifying thyroid nodule tissue.

Abstract: Examples are disclosed that relate to computing device input systems. In one example, a computing device input system comprises an input configured to receive a connection to a control device connector comprising a plurality of conductors. The computing device input system further comprises circuitry configured to determine a presence, type, and state of a control device in communication with the control device connector based on analog voltages received from the control device connector.

Abstract: An accessory apparatus includes a housing, an apparatus connector, a plurality of auxiliary accessory interfaces, and an internal microcontroller. The connection connector is configured to mate with a corresponding accessory connector of a physical controller to electrically connect the internal microcontroller to the physical controller. Each auxiliary accessory connector is configured to enable a separate auxiliary user input device to operatively connect to the accessory apparatus and electrically connect with the internal microcontroller. The internal microcontroller is configured to: (1) receive an input control signal from an auxiliary user input device operatively connected to an auxiliary accessory connector of the plurality of auxiliary accessory connectors, (2) map the input control signal to a mapped control signal corresponding to a physical control of the physical controller, and (3) send the mapped control signal to the physical controller via the apparatus connector.