Abstract: A system and method of automatic product attribute recognition receive training images having bounding boxes associated with one or more products in the training images, receive attribute values for each of the one or more products in the training images, and train a first convolutional neural network (CNN) model to generate bounding boxes for and identify each of the one or more products with the training images until the accuracy of the first CNN model is above a first predetermined threshold. The system and method further train a second CNN model for each of the products associated with the cropped images until the second CNN generates attribute values for the one or more attributes with an accuracy above a second predetermined threshold, and automatically recognize the one or more attributes for a new product image by presenting the product image to the first and second CNN models.

Abstract: A system (100) and a method (200) for exchanging private data among nodes (102a, 102b, . . . , 102n) in a blockchain network (102). The method (200) comprises defining one or more groups of nodes; implementing a structured network topology with one or more network zones for establishing connections between the nodes; assigning a topology identification number to each of the nodes, such that the topology identification numbers of the nodes in each of the network zones are in a series; defining a limit for each of the network zones based on the topology identification numbers of the nodes therein; and exchanging private data between the nodes in the blockchain network in consideration of the defined limits for the corresponding network zones thereof.

Abstract: A system and method of automatic product attribute recognition receive training images having bounding boxes associated with one or more products in the training images, receive attribute values for each of the one or more products in the training images, and train a first convolutional neural network (CNN) model to generate bounding boxes for and identify each of the one or more products with the training images until the accuracy of the first CNN model is above a first predetermined threshold. The system and method further train a second CNN model for each of the products associated with the cropped images until the second CNN generates attribute values for the one or more attributes with an accuracy above a second predetermined threshold, and automatically recognize the one or more attributes for a new product image by presenting the product image to the first and second CNN models.

Abstract: A system and method of automatic product attribute recognition receive training images having bounding boxes associated with one or more products in the training images, receive attribute values for each of the one or more products in the training images, and train a first convolutional neural network (CNN) model to generate bounding boxes for and identify each of the one or more products with the training images until the accuracy of the first CNN model is above a first predetermined threshold. The system and method further train a second CNN model for each of the products associated with the cropped images until the second CNN generates attribute values for the one or more attributes with an accuracy above a second predetermined threshold, and automatically recognize the one or more attributes for a new product image by presenting the product image to the first and second CNN models.

Abstract: A system and method of automatic product attribute recognition receive training images having bounding boxes associated with one or more products in the training images, receive attribute values for each of the one or more products in the training images, and train a first convolutional neural network (CNN) model to generate bounding boxes for and identify each of the one or more products with the training images until the accuracy of the first CNN model is above a first predetermined threshold. The system and method further train a second CNN model for each of the products associated with the cropped images until the second CNN generates attribute values for the one or more attributes with an accuracy above a second predetermined threshold, and automatically recognize the one or more attributes for a new product image by presenting the product image to the first and second CNN models.

Abstract: Embodiments herein disclose a method for MBS switching and continuity in a 5G wireless network. The method includes establishing, by a gNodeB, a MBS session with a UE. Further, the method includes providing, by the gNodeB, MBS to the UE during the MBS session. Further, the method includes detecting, by the gNodeB, that the at least one MBS is going to get suspended or deactivated. Further, the method includes sending, by the gNodeB, an indication to the UE indicating the at least one MBS going to get suspended or deactivated prior to suspension or deactivation of the at least one MBS at the UE.

Abstract: Disclosed herein are systems and methods for identifying service centers. One of such methods may comprise operating at least one processor to: receive telematics data originating from a plurality of telematics devices installed in a plurality of vehicles; determine for each of the plurality of vehicles, using the telematics data, when a vehicle maintenance event has occurred by identifying at least a binary vehicle status indicator change, time-logged telematics data that meets a predetermined condition, an indication that a vehicle diagnosis has occurred, or a combination thereof; determine, using the telematics data, a location of each of the plurality of vehicles at a time at which the vehicle maintenance event occurred; and identify one or more service centers by: applying to a plurality of maintenance event locations a clustering model, and classifying one or more clusters of maintenance event locations as a service center.

Abstract: Disclosed herein are systems and methods for estimating remaining useful life of vehicle engine oil.

Abstract: A system and a method for writing and retrieval of data in a blockchain with a plurality of nodes adapted to manage multiple assets is provided. The method includes identifying nodes that are offline and determining by the nodes itself whether it is online or offline, at a current instant; storing, by the nodes that are offline, corresponding transaction data related to managed assets, locally therein; determining if any of the offline nodes at a previous instant is back online at the current instant; receiving transaction data from such nodes that are back online; and recording the received transaction data to a current block in the blockchain, if there has been no transaction data for the asset recorded in any of the blocks committed between the node was last identified to be offline and the current instant.

Abstract: A system and method are disclosed for aggregating product trends. Embodiments include receiving an initial set of images from one or more data feeds by a trend aggregation system comprising a server, identifying one or more attribute values from the initial set of images, and quantifying a social affinity score of one or more items of an inventory of a supply chain entity based on recentness, relevance, and similarities of the identified one or more attribute values to an attribute value of a potential product for a product assortment. Embodiments may further include receiving an image of at least one additional item, identifying a product attribute from the image, and assigning an attribute value to the at least one additional item based, at least in part, on the identified attribute value from the image of the least one additional item and the attribute values of the at least two items.

Abstract: A system and method of automatic product attribute recognition receive training images having bounding boxes associated with one or more products in the training images, receive attribute values for each of the one or more products in the training images, and train a first convolutional neural network (CNN) model to generate bounding boxes for and identify each of the one or more products with the training images until the accuracy of the first CNN model is above a first predetermined threshold. The system and method further train a second CNN model for each of the products associated with the cropped images until the second CNN generates attribute values for the one or more attributes with an accuracy above a second predetermined threshold, and automatically recognize the one or more attributes for a new product image by presenting the product image to the first and second CNN models.

Abstract: Systems and methods for detecting and removing erroneous tire pressure values are provided. The method involves operating at least one processor to: receive tire pressure data comprising a time series of tire pressure values; simplify the tire pressure data by removing at least some of the tire pressure values that satisfy a predetermined interpolation error threshold; detect, using a trained machine learning classifier, at least one erroneous tire pressure value in the tire pressure data, each erroneous tire pressure value being detected based on the erroneous tire pressure value and a plurality of lagging tire pressure values consecutively trailing the erroneous tire pressure value; clean the tire pressure data by removing the at least one erroneous tire pressure value; and transmit the tire pressure data to a fleet management system, whereby the at least one erroneous tire pressure value is not transmitted to the fleet management system.

Abstract: Embodiments herein disclose a method for MBS switching and continuity in a 5G wireless network. The method includes establishing, by a gNodeB, a MBS session with a UE. Further, the method includes providing, by the gNodeB, MBS to the UE during the MBS session. Further, the method includes detecting, by the gNodeB, that the at least one MBS is going to get suspended or deactivated. Further, the method includes sending, by the gNodeB, an indication to the UE indicating the at least one MBS going to get suspended or deactivated prior to suspension or deactivation of the at least one MBS at the UE.

Abstract: An emergency call location manager on a computer can use a settings file stored on the computer to determine a physical address that corresponds to a network address associated with the computer, such that the emergency call location manager can provide the physical address as dispatchable location information when a softphone application on the computer makes an emergency call. Task schedulers on computers can automatically execute tasks to download the settings file from a remote repository before the emergency call location manager begins executing on the computers, and to upload changes to the settings file made locally on the computers to the remote repository. Accordingly, the tasks can cause the emergency call location manager on a computer to use the most recent version of the settings file, even if the settings file was originally created on a different computer or a different computer made changes to the settings file.

Abstract: Systems and methods for detecting and removing erroneous tire pressure values are provided. The method involves operating at least one processor to: receive tire pressure data comprising a time series of tire pressure values; simplify the tire pressure data by removing at least some of the tire pressure values that satisfy a predetermined interpolation error threshold; detect, using a trained machine learning classifier, at least one erroneous tire pressure value in the tire pressure data, each erroneous tire pressure value being detected based on the erroneous tire pressure value and a plurality of lagging tire pressure values consecutively trailing the erroneous tire pressure value; clean the tire pressure data by removing the at least one erroneous tire pressure value; and transmit the tire pressure data to a fleet management system, whereby the at least one erroneous tire pressure value is not transmitted to the fleet management system.

Abstract: Systems and methods for detecting and removing erroneous tire pressure values are provided. The method involves operating at least one processor to: receive tire pressure data comprising a time series of tire pressure values; simplify the tire pressure data by removing at least some of the tire pressure values that satisfy a predetermined interpolation error threshold; detect, using a trained machine learning classifier, at least one erroneous tire pressure value in the tire pressure data, each erroneous tire pressure value being detected based on the erroneous tire pressure value and a plurality of lagging tire pressure values consecutively trailing the erroneous tire pressure value; clean the tire pressure data by removing the at least one erroneous tire pressure value; and transmit the tire pressure data to a fleet management system, whereby the at least one erroneous tire pressure value is not transmitted to the fleet management system.

Abstract: An electronic device package may include a substrate which may include a dielectric medium. A transition cavity may be defined by the dielectric medium, the transition cavity may be sized and shaped for mechanical coupling with a waveguide structure. The electronic device package may also include a first antenna and a second antenna, the first antenna and the second antenna may be positioned to excite the transition cavity, the first antenna may be fed by a first transmission line and the second antenna may be fed by a second transmission line. The electronic device package may also include an electromagnetic probe which may be electrically coupled with the transition cavity. The electromagnetic probe may interact with an electromagnetic field in the transition cavity and may provide an electrical signal representative of the electromagnetic field in the transition cavity. The first transmission line may include a first planar conductor.

Abstract: An electronic device package may include an integrated circuit die mounted on a package substrate, the package substrate may include a dielectric medium, the integrated circuit die may include respective electrical interconnects located between the integrated circuit die and the dielectric medium. The electronic device package may also include a first transmission line which may be embedded within the dielectric medium, the first transmission line may be fed by the integrated circuit die from at least one of the respective electrical interconnects and through a corresponding first via structure in the dielectric medium. The integrated circuit die may one or more of include or be attached to a conductive plane that extends to overlap with a location of the first via structure in the dielectric medium. The conductive plane may be connected to a first return structure of the first transmission line by one or more conductive pillars.

Abstract: An integrated device package may include a glass interposer having one or more conductive vias extending through the glass interposer from a top side of the glass interposer to a bottom side of the glass interposer, the bottom side of the glass interposer comprising one or more contact pads. The device package may include an integrated device die mounted and electrically connected to the top side of the glass interposer. The device package may also include an encapsulating material disposed over at least a side surface of the glass interposer, a portion of the top surface of the glass interposer, and a side surface of the integrated device die.

Abstract: The innovation disclosed and claimed herein, in one aspect thereof, comprises systems and methods of funds transfer authentication. The authentication can include receiving a transfer request to transfer funds from a payer to a payee, the transfer request identifies the payer, the payee, and an amount to transfer. Customer information associated respectively with the payer and the payee is identified. The payer and the payee are authenticated. A communication session is opened between the payer and payee. A time-limited payee OTP is generated and sent to the payee. The payee OTP is received within a first time limit. The funds are transferred from the payer to the payee based on receiving a matching payee OTP.