Abstract: A method includes obtaining image data, identifying a machine learning-compressible (ML-compressible) portion of the image data, and determining a location of the ML-compressible portion within the image data. The method also includes selecting, from a plurality of ML compression models, an ML compression model for the ML-compressible portion based on an image content thereof, and generating, based on the ML-compressible portion and by the ML compression model, an ML-compressed representation of the ML-compressible portion. The method further includes generating a compressed image data file that includes the ML-compressed representation and the location of the ML-compressible portion, and outputting the compressed image data file. The compressed image data file is configured to cause an ML decompression model corresponding to the ML compression model to generate a reconstruction of the ML-compressible portion of the image data based on the ML-compressed representation.

Abstract: A system for producing and packaging fluid may comprise a water distillation device. The system may further comprise a mixing circuit coupled to an output of the water distillation device and including a source of concentrate. The mixing circuit may include a plurality of flow controllers configured to adjust the flow of fluid through the mixing circuit so as to generate a preset fluid. The system may further comprise an enclosure including an antechamber and a packaging compartment. The system may further comprise a reservoir dispenser in the packaging compartment having a feed plate and a housing block. The reservoir dispenser may include a bias member which urges the feed plate toward the housing block. The system may further comprise a filling station in the packaging compartment including a filling nozzle coupled to the mixing circuit. The system may further comprise a sealing station in the packaging compartment having a ram and a sealing member dispenser.

Abstract: Aspects of the present disclosure provide systems and methods for detecting memorable events of a user, storing those memorable events in a user profile, and leveraging those events to the user. In so doing, memorable events with positive emotional weight may be recalled to encourage delightful experiences. Alternatively or additionally, memorable events with negative emotional weight may be recalled to increase user engagement for conversational systems.

Abstract: Techniques of wireless charging involve a charging base configured to predict when the user is about to charge a device based on a state of the device. For example, the state of the device may be defined by inertial measurement units (IMUs). The charging base then activates in response to predicting that the user intends to charge the device. In some implementations, the prediction is based on a machine learning (ML) engine trained to classify a movement of a device by a user as being about to charge or not being about to charge. If a movement of a device is classified as being about to charge, the charging base activates before the device is detected by the charging base.

Abstract: To provide navigation directions for a scenic route, a machine learning model is trained using (i) characteristics of road segments that have been assigned a scenic metric and (ii) the scenic metrics for the road segments. In response to a request for navigation directions, a set of candidate routes for navigating from the starting location to the destination location is identified. Then for each road segment within each candidate route, characteristics of the road segment are applied to the machine learning model to generate the scenic metric for the road segment. A route is then selected from the set of candidate routes based at least in part on the scenic metrics of the road segments within the route. A set of navigation directions is provided for presentation on a client device for navigating from the starting location to the destination location via the selected route.

Abstract: Techniques of wireless charging involve a charging base configured to predict when the user is about to charge a device based on a state of the device. For example, the state of the device may be defined by inertial measurement units (IMUs). The charging base then activates in response to predicting that the user intends to charge the device. In some implementations, the prediction is based on a machine learning (ML) engine trained to classify a movement of a device by a user as being about to charge or not being about to charge. If a movement of a device is classified as being about to charge, the charging base activates before the device is detected by the charging base.

Abstract: The present disclosure is directed towards systems and methods for receiving environmental data from device sensors. A computing system stores environmental data in an environmental feature database at the computing system for a plurality of geographic locations. The computing system receives, from one or more remote systems, data indicating one or more environmental features for a particular geographic location. The computing system accesses stored environmental data for the particular geographic location to determine whether the environmental features are included in the environmental feature database. In response to determining that the environmental features are included in the environmental feature database, the operations further comprise, updates a confidence value associated with the environmental features.

Abstract: To provide navigation directions for a scenic route, a machine learning model is trained using (i) characteristics of road segments that have been assigned a scenic metric and (ii) the scenic metrics for the road segments. In response to a request for navigation directions, a set of candidate routes for navigating from the starting location to the destination location is identified. Then for each road segment within each candidate route, characteristics of the road segment are applied to the machine learning model to generate the scenic metric for the road segment. A route is then selected from the set of candidate routes based at least in part on the scenic metrics of the road segments within the route. A set of navigation directions is provided for presentation on a client device for navigating from the starting location to the destination location via the selected route.

Abstract: Examples of the present disclosure describe systems and methods of creating a conversational chat bot of a specific person. In aspects, social data (e.g., images, voice data, social media posts, electronic messages, written letters, etc.) about the specific person may be accessed. The social data may be used to create or modify a special index in the theme of the specific person's personality. The special index may be used to train a chat bot to converse in the personality of the specific person. During such conversations, one or more conversational data stores and/or APIs may be used to reply to user dialogue and/or questions for which the social data does not provide data. In some aspects, a 2D or 3D model of a specific person may be generated using images, depth information, and/or video data associated with the specific person.

Abstract: Systems and methods for personality consistent automated chatting are provided. The systems and methods track the emotional state of the chat bot, track entity related dispositions of the chat bot and/or track the closeness of the relationship between the chat bot and a user. Accordingly, the systems and methods are able to provide consistent personality and/or points of view by generating replies based on the tracked chat bot emotional state, chat bot entity dispositions, and/or the relationship closeness between the chat bot and the user.

Abstract: Aspects of the present disclosure provide systems and methods for detecting memorable events of a user, storing those memorable events in a user profile, and leveraging those events to the user. In so doing, memorable events with positive emotional weight may be recalled to encourage delightful experiences. Alternatively or additionally, memorable events with negative emotional weight may be recalled to increase user engagement for conversational systems.

Abstract: Systems and methods for personality consistent automated chatting are provided. The systems and methods track the emotional state of the chat bot, track entity related dispositions of the chat bot and/or track the closeness of the relationship between the chat bot and a user. Accordingly, the systems and methods are able to provide consistent personality and/or points of view by generating replies based on the tracked chat bot emotional state, chat bot entity dispositions, and/or the relationship closeness between the chat bot and the user.

Abstract: Examples of the present disclosure describe systems and methods of creating a conversational chat bot of a specific person. In aspects, social data (e.g., images, voice data, social media posts, electronic messages, written letters, etc.) about the specific person may be accessed. The social data may be used to create or modify a special index in the theme of the specific person's personality. The special index may be used to train a chat bot to converse in the personality of the specific person. During such conversations, one or more conversational data stores and/or APIs may be used to reply to user dialogue and/or questions for which the social data does not provide data. In some aspects, a 2D or 3D model of a specific person may be generated using images, depth information, and/or video data associated with the specific person.

Abstract: Described are computer-based methods and apparatuses, including computer program products, for managing insurance plans. Multiple bids are received from multiple reinsurance companies for the price of a unit of an insurance product (e.g., a fixed annuity paying ten dollars a month starting at the age of sixty five). The unit of the insurance product is associated with a set of one or more risk factors (e.g., age, range of ages). A bid is selected from the bids based on a relationship between the bids (e.g., lowest bid, lowest weighted bid). The selected bid is associated with a selected reinsurance company. Units of the insurance product associated with the selected bid are offered to participants associated with the set of one or more risk factors.

Abstract: Described are computer-based methods and apparatuses, including computer program products, for managing insurance plans. Multiple bids are received from multiple reinsurance companies for the price of a unit of an insurance product (e.g., a fixed annuity paying ten dollars a month starting at the age of sixty five). The unit of the insurance product is associated with a set of one or more risk factors (e.g., age, range of ages). A bid is selected from the bids based on a relationship between the bids (e.g., lowest bid, lowest weighted bid). The selected bid is associated with a selected reinsurance company. Units of the insurance product associated with the selected bid are offered to participants associated with the set of one or more risk factors.