Abstract: A method for creating a student model from a teacher model for knowledge distillation. The method may include: providing the teacher model trained on a first training dataset; generating candidate student models, wherein each of the candidate student models is a model having a unique permutation of layers derived by randomly selecting one or more layers of the plurality of layers of the teacher model for removing; generating a second training dataset; for each of the candidate student models: providing the second training dataset as inputs; recording outputs generated; and based on the recorded outputs, evaluating a performance according to a predetermined model evaluation criterion; determining which of the candidate student models performed best among the candidate student models based on the predetermined model evaluation criterion; identifying a preferred candidate student model.

Abstract: A sucker rod guide is used on a sucker rod within production tubing. The guide includes a sleeve having multiple vanes. The sleeve is designed to fit onto the sucker rod and includes ends, an outer surface, and a particular length and diameter along its axis. The vanes are evenly distributed about the sleeve and protrude outwardly to a defined diameter. Each vane exhibits symmetry along longitudinal and lateral axes, with distinct characteristics. Sidewalls at each end of each vane diverge from one another and extend to a midpoint of symmetry. Angled faces at each end of each vane extend at an angle from the sleeve's surface, and each vane has an external face that has a diamond shape. The guide is configured to reduce turbulence adjacent the sucker rod near the edge of the guide.

Abstract: Systems, methods, and products described herein include a server, which receives a request from a user's electronic client device. The server may query web pages and determine if the web pages comply with the compliance rules. When the web pages do not comply with the compliance rules, the server may generate an analysis report on the web pages and send the analysis report to the user. After the user reviews the report and chooses the trouble web pages, the server sends review alerts to the points of contact, such as the product owners and other stakeholders. The product owner may more easily be able to take proper actions and better manage the websites by refreshing noncompliant or incorrect content. In this way, the systems and methods allow product owners and users to manage and/or curate the content of web pages more systematically and automatically.

Abstract: A system and techniques are described herein for providing authentication. The technique includes registering user authentication data such as biometrics data with a communication device. The authentication data is linked to an account or service provider, and is used to verify the identity of the user when accessing the account. The communication device may obtain a public/private key pair, for which the public key may be stored on a secure remote server. When the user attempts to access the account or service provider, the user may provide the authentication data to authenticate the user to the communication device. Thereafter, the communication device may sign an authentication indicator using the private key and send the authentication indicator to the secure remote server. Upon verification of the signature using the public key, the secure remote server may grant access to the user, for example, by releasing a token.

Abstract: A modular transaction card assembly includes a card frame having the traditional dimensions of a credit card, and a transaction card that is smaller than a traditional card and that fits into a receptacle of the card frame. The card frame may include a window in which contact pads of the transaction card are positioned and flush with the surface of the card frame when the transaction card is secured in the receptacle. The card frame and transaction card assembled together may form a uniform assembly that appears and functions as a traditional transaction card (e.g., credit card).

Abstract: A call center data transfer system includes one or more processors that are configured to receive call data related to information provided by a caller during a first portion of a call with a first customer service representative. The one or more processors are also configured to assign an identifier to the call in response to receipt of an indication that the call should be transferred from the first customer service representative, associate the identifier with the call data in one or more databases, receive the identifier as an input during a second portion of the call with a second customer service representative, and use the identifier to retrieve the call data from the one or more databases. The one or more processors are further configured to display the call data via a display for visualization by the second customer service representative.

Abstract: Wellbore logging apparatus include a positioning mechanism that couples one or more well logging devices included in a well logging tool to a downhole tractor. The positioning tool allows the logging tools to be moved at a different rate of travel through a wellbore compared to the rate of travel of the downhole tractor through the wellbore.

Abstract: Aspects of the subject disclosure may include, for example, identifying a communication device communicatively coupled to a serving base station, requesting a device type associated with the communication device from the serving base station, and determining the communication device comprises an unmanned aerial vehicle based on the device type. Further embodiments can include obtaining a geolocation of the communication device based on the communication device comprising the unmanned aerial vehicle, generating a customized radio resource control (RRC) reconfiguration message, which includes a reportCGI object that involves setting a discontinuous sleep state timer, which stops the data transmission and reception of the communication device with the serving base station. During this discontinuous sleep state timer, the communication device disconnects to its serving base station to read and decode the ECGI of base station associated with neighboring cells. Other embodiments are disclosed.

Abstract: A device, system, and method for maintaining temperature control of a distribution item, or of the contents of a distribution item, as the distribution item moves through the distribution network. A container housing an item can include a cooling unit, a heating unit, or both, and control circuitry including a temperature sensor. The control circuitry activates the cooling unit or heating unit as required to maintain the item at a desired temperature or within a desired temperature range.

Abstract: Loading, securing, transporting, and/or depositing objects as well as systems, methods, and apparatuses for the same, including those that operate in automated or semi-automated fashion, are disclosed herein. In one embodiment, a mobile robotic platform is provided. The mobile robotic platform is designed to receive, transport, and deposit objects at desired destinations. In another embodiment, a flat transport surface, e.g., a pallet, is provided. The flat transport surface is designed to facilitate efficient loading, transport, and deposit of objects, e.g., between automated or semi-automated handling systems, in one aspect. In another embodiment, a system for handling objects is provided. The system is designed to support automated or semi-automated loading and/or unloading of objects, e.g., onto or from a mobile robotic platform, and may be implemented in a vehicle, in one aspect.

Abstract: A method for acoustic noise source detection. The method may comprise disposing an acoustic logging tool into a wellbore, performing an acoustic logging operations in the wellbore with the acoustic logging tool, forming a data set form the acoustic logging operation, and performing a cross-correlation function between pairs of hydrophones using the data set. The method may further comprise construction a cost function using at least in part the cross-correlation function to find a noise source distribution and identifying a location of the acoustic noise source by inverting the noise source distribution.

Abstract: The invention is a security system providing domain name authentication for intrusion and malware prevention. The system is configured to analyze domain names, specifically analyze network metadata associated with said domain names, and further identify domain names attempting to impersonate or spoof domain names associated with a trusted entity or party.

Abstract: Accuracy of transcript is of foremost importance in Automatic Speech Recognition (ASR). State of the art system mostly rely on spelling correction based contextual improvement in ASR, which is generally a static vocabulary based biasing approach. Embodiments of the present disclosure provide a method and system for visual context aware ASR. The method provides biasing using shallow fusion biasing approach with a modified beam search decoding technique, which introduces a non-greedy pruning strategy to allow biasing at the sub-word level. The biasing algorithm brings in the visual context of the robot to the speech recognizer based on a dynamic biasing vocabulary, improving the transcription accuracy. The dynamic biasing vocabulary, comprising objects in a current environment accompanied by their self and relational attributes, is generated using a bias prediction network that explicitly adds label to objects, which are detected and captioned via a state of the art dense image captioning network.

Abstract: A transaction card construction and a method for making transaction cards provides increased security for transaction card magnetic strips. The transaction card construction includes a card inlay and a clear card body. The card inlay is formed via a lamination press process with the magnetic strip attached to a back surface of the card inlay. The card body may have a window through which a data storage element may be exposed for accessing, such as by a magnetic stripe reader or EMV chip reader. The card body may be formed by adhering the card inlay to the clear card body.

Abstract: A method may include receiving data related to a plurality of items and processing the data using a machine learning model. The machine learning model may have been trained to output a score for each of the plurality of items based on one or more target variables and to process the data using a grouped linear regression for groups of items based on sub-divisions of the groups. The method may include storing the output in a data store. Each entry in the data store may include at least an item identifier for an item, a group name, and the score. The method may include receiving search criteria for a search and identifying a set of search results in a group of items. The method may include determining an order of the set of search results and outputting the set of search results.

Abstract: Auditing data containing sensitive data are stored in a data structure comprising data objects. Each data object comprises one or more pairs of a name and a value. Pairs that are flagged or identified as containing sensitive data are partially encrypted; the value is encrypted using an asymmetric key and the name corresponding to the encrypted value remains unencrypted. Some pairs that are not flagged or identified as containing sensitive data are left unencrypted. Unencrypted data may be stored in the partially encrypted auditing data as plain text. The auditing data may be analyzed to generate business metrics and identify application errors. The auditing data may also be queried, and data objects containing unencrypted pairs and/or partially encrypted pairs may be returned based on matching unencrypted names and/or values to the data query.

Abstract: Systems and methods for authentication may include a first device having an association with a first account, including a memory containing one or more applets, a counter value, and transmission data, a communication interface, and one or more processors in communication with the memory and communication interface. The first device may create a cryptogram based on the counter value, wherein the cryptogram includes the counter value and the transmission data. The first device may transmit, after entry of the communication interface into a communication field, the cryptogram, and update, after transmission of the cryptogram, the counter value. The first device may receive, via the communication interface, one or more encrypted keys and one or more parameters. The first device may decrypt the one or more encrypted keys and, after decryption of the one or more encrypted keys, switch an association from the first account to a second account.

Abstract: A multi-bore junction assembly to be positioned in a multilateral well having a main wellbore and a lateral wellbore, wherein the multilateral well has an up-hole end and a downhole end, the multi-bore junction assembly comprises a deflector subassembly to be positioned at a junction between the main wellbore and the lateral wellbore. The deflector subassembly comprises a deflector sub that comprises, a first deflector bore at an up-hole end of the deflector sub, a second deflector bore at a downhole end of the deflector sub, wherein the second deflector bore is radially offset from the first deflector bore, a seal stinger to project from the second deflector bore, and a slot sub. Additionally, the multi-bore junction assembly comprises an outer sleeve positioned around the deflector subassembly, the outer sleeve having at least one pin, wherein the slot sub is to interact with the at least one pin.

Abstract: A therapy system includes a patient support apparatus and a pneumatic therapy device that is coupleable to the patient support apparatus. The therapy device may receive power and air flow from the patient support apparatus.

Abstract: Some implementations relate to a computer-implemented method for creating digital acoustic sensing (DAS) related training data for a learning machine. The method may include moving a first optimal microseismic event location to a first perturbed microseismic event location in each of a plurality of first images. The method also may include modifying first shear waves and first compressional waves in each of the first images based on one or more signal travel times between the first perturbed microseismic event location and a fiber optic cable to form a plurality of training images configured to train a learning machine.