Open Access. Powered by Scholars. Published by Universities.®
Articles 3121 - 3150 of 4090
Full-Text Articles in Entire DC Network
Systems And Methods For Machine Learning Brand Identification Models, Roy Palacios, Pallavi Satyan
Systems And Methods For Machine Learning Brand Identification Models, Roy Palacios, Pallavi Satyan
Defensive Publications Series
A high precision machine learning model trained to identify brands is described. One or more machine-learned models are trained to create signals for top product categories at the brand/merchant level and brand signals from a custom Brand Ownership Service based on external digital brand signals such as trademark data from the USPTO, the EUIPO, and/or other trademark registries.
Lens Carrier And Waveguide Subframe For Near-Eye Device, Lloyd Holland, Joshua Moore, Jaehong Choi
Lens Carrier And Waveguide Subframe For Near-Eye Device, Lloyd Holland, Joshua Moore, Jaehong Choi
Defensive Publications Series
This disclosure describes a lens stack (lens module) for a near-eye device that includes a lens carrier and waveguide subframe. The lens carrier and waveguide subframe fully or partially encircles and houses the waveguide and enables precision placement of the waveguide in-coupling grating with respect to controlled datums that interface with display modules. The lens carrier is made of a high strength material and provides structural support to the waveguide. The material is opaque and helps to minimize stray light in the display system by reducing edge glow from waveguide. The lens carrier and waveguide subframe additionally provide a locating …
Intelligently Leveraging Mental Wellness State In Cognitive Collaboration Workflows, Bhavik Shah, Chidambaram Arunachalam, John Groetzinger
Intelligently Leveraging Mental Wellness State In Cognitive Collaboration Workflows, Bhavik Shah, Chidambaram Arunachalam, John Groetzinger
Defensive Publications Series
A person's mental wellness plays a critical role in determining how efficiently a complex task that requires cognitive abilities is successfully completed. However, current solutions do not offer a mechanism for detecting an individual’s mental wellness in a real-time and proactive manner. To address that lack, techniques are presented herein that leverage a range of information including electroencephalography (EEG) signals (e.g., as captured from an EEG headset); real-time health parameters such as heart rate, blood pressure, etc. (e.g., as captured from health monitoring devices such as wearable physical fitness monitors); etc. Such information may be processed by an online communication …
Using Cross-User Understanding To Develop Better User Embeddings, Nithya Kota, Maneesh Dewan, Ganesh Mallya, Jatin Chhugani
Using Cross-User Understanding To Develop Better User Embeddings, Nithya Kota, Maneesh Dewan, Ganesh Mallya, Jatin Chhugani
Defensive Publications Series
Digital payment apps, e.g., available via smartphones, can make multiple payment instruments available to a user; however, a user may not link all their cards to a payment app. Some payment apps include analytics features based on data from cards linked to the app. However, such data can give an incomplete financial picture of the user. This disclosure describes techniques to develop a more complete representation of a user that has provided only a partial view of their finances. Two embedding vector spaces are created, one trained over users with incomplete financial profiles and another trained over users with complete …
Authentication And Key Management Of Iot Applications With Extended Wifi Authentication (Wifi Akma), Rajesh I V, Ram Mohan R
Authentication And Key Management Of Iot Applications With Extended Wifi Authentication (Wifi Akma), Rajesh I V, Ram Mohan R
Defensive Publications Series
Internet of things (IoT) devices frequently apply insufficient authentication mechanisms with their application servers due to the constrained nature of such devices. For example, most IoT devices lack the resources that are necessary to store usernames and passwords, certificates, and keys in a secured manner. The challenge that was described above is solved in a 3rd Generation Partnership Project (3GPP) fifth-generation (5G) wireless environment through the Authentication and Key Management for Applications (AKMA) initiative. However, there is no AKMA-equivalent facility within a WiFi environment. Accordingly, techniques are presented herein that extend the WiFi authentication process to support application server authentication …
Using Federated Learning To Enhance Image Transformations On Learning-Based Techniques, Hp Inc
Using Federated Learning To Enhance Image Transformations On Learning-Based Techniques, Hp Inc
Defensive Publications Series
This disclosure defines a new strategy to enhance image transformation approaches for image scale-up and super-resolution (SR) using a federated learning architecture to augment image resolution. The main idea relies on using FL which allows the tunning of learning algorithms in a decentralized way without the necessity of any sensitive data leaving the client’s device. Hence, it is possible to train models to adjust image resolution directly on customer devices, using locally available information to learn the patterns most relevant to the desired task in an self-supervised and privacy-guaranteed way. FL can optimize models indirectly by using the info from …
A Workflow For Document Template Extraction Using Machine Learning, Hp Inc
A Workflow For Document Template Extraction Using Machine Learning, Hp Inc
Defensive Publications Series
This disclosure aims to allow the user the possibility to extract the template of a document using only a picture of it. Our method can be described as follows: 1. The document image is identified in the scene (when mobile scanning, for instance); 2. The document image is cropped from the scene; 3. The document image is segmented into its different regions: e.g., title, image, graphic, content text; 4. For each region, an algorithm (e.g., a ML model) will be used to extract the region template features and an OCR engine will extract the text from the region; 5. The …
Improved Bridging In Fff, Paul Kuiper, Siert Wijnia
Improved Bridging In Fff, Paul Kuiper, Siert Wijnia
Defensive Publications Series
This publication relates to a printing strategy that increases the visual quality and dimensional accuracy of “bridges” in 3D printed objects using FFF.
An Electric Assisted Turbo With Centrifugal Clutch And Axial Flux Motor, Bin Zhu, King Jin, Huan Wang
An Electric Assisted Turbo With Centrifugal Clutch And Axial Flux Motor, Bin Zhu, King Jin, Huan Wang
Defensive Publications Series
A traditional electric turbocharger (E-turbo) uses a radial flux motor coupled to the turbocharger to provide additional power to the drive the compressor at low speeds. This configuration is not compact and lacks satisfying transient response when motor is powered off because of high inertia of the wheel rotor system. This article proposes a novel E-turbo configuration that uses an axial flux motor and a centrifugal clutch so that the E-turbo is more compact with more flexible connection between motor and wheel rotor system.
Unitary Ported Shroud For Compressor Housing For Flow Stability And Noise Reduction, Mohan Javvadi
Unitary Ported Shroud For Compressor Housing For Flow Stability And Noise Reduction, Mohan Javvadi
Defensive Publications Series
Ported Shroud for a turbocharger allows air from downstream of compressor to recirculate back to upstream in order to increase flow stability and avoid surge. However, this results in an increase in noise produced by turbocharger. Noise suppressors that are typically added at the inlet to address this problem make inlet assembly complex and increase the cost. This article proposes a novel design of ported shroud that also functions as noise suppressor.
Camera Focus Adjustment Using Depth Estimated Via Ultra-Wideband (Uwb) Handshake, D Shin
Camera Focus Adjustment Using Depth Estimated Via Ultra-Wideband (Uwb) Handshake, D Shin
Defensive Publications Series
Many video conferencing applications support portrait mode where the participant is in focus and their surroundings are blurred. Portrait mode based on computer vision techniques can sometimes be unsatisfactory due to difficulties in separating the user and the background. This disclosure describes techniques that use existing ultra-wideband (UWB) hardware on commodity devices to perform crisp and more accurate segmentation. A UWB handshake protocol is utilized to estimate the distance and angle between the camera and another device that the conference participant is wearing. The estimate is used to automatically adjust camera focal length to focus on the conference participant while …
Personalized Shopping Via Conversational User Interface, Abhishek Yadav, Amit Handa
Personalized Shopping Via Conversational User Interface, Abhishek Yadav, Amit Handa
Defensive Publications Series
Online and offline shopping activity often requires shoppers to spend substantial amounts of time researching, finding, and refining their product search to identify the right product that meets their requirements. Some of this time may be spent in searching and filtering results on general-purpose search engines or merchant websites, neither of which are optimized for shopping related research. This disclosure describes techniques that enable shopping via a rich, multimedia conversational interface. The techniques provide an online shopping experience that is simple, uncluttered, and does not overwhelm the user. The user is provided guidance throughout their shopping journey.
Behavior Of A Multi-Planetary Search Engine, Ashish Duggal
Behavior Of A Multi-Planetary Search Engine, Ashish Duggal
Defensive Publications Series
The present disclosure describes computer-implemented systems and methods for providing search results to users located in remote areas utilizing local hardware and locally cached data to return search results at little or no cost and providing a greater search breadth tier that utilizes remote hardware and data with an associated cost. User profiles associated with the users located in the remote areas may drive an aggregate, anonymized and frequency adjust model to determine subjects of interest to the plurality of users.
Automating Network Device Configuration Template Discovery, Derek Engi, Gonzalo Salgueiro, Ravi Garg, Bradley Wise
Automating Network Device Configuration Template Discovery, Derek Engi, Gonzalo Salgueiro, Ravi Garg, Bradley Wise
Defensive Publications Series
Establishing network device "golden" configuration templates typically relies on largely manual efforts and a dialogue between network consulting engineers and customers. Techniques are presented herein that streamline the process of discovering and baselining network standards by examining underlying themes and relationships between micro-templates in customer environments. Aspects of the presented techniques employ a Masked-Language Model (MLM), in a way that an MLM was not necessarily intended, to detect locally significant attributes in network device configurations for the express purpose of identifying structures that are common across a set of devices. Under further aspects of the presented techniques, the creation of …
Predictive Ev Mode Drop In Phev To Prevent Mileage Collapse, Axel Unger
Predictive Ev Mode Drop In Phev To Prevent Mileage Collapse, Axel Unger
Defensive Publications Series
No abstract provided.
Method To Simplify Detailed Testing In Distributed Microservice Architectures, Anonymous
Method To Simplify Detailed Testing In Distributed Microservice Architectures, Anonymous
Defensive Publications Series
This disclosure describes a mechanism in which a user action(s) can be executed, and the internals of the system recorded when the actions are manually run, so that the system behaviour through to completion of the action(s) can be captured, analyzed, modified, shared, and compared against in subsequent iterations.
Method For Minimizing Over Booking Of Resources In Rhp Enabled Control Plane Networks, Anonymous
Method For Minimizing Over Booking Of Resources In Rhp Enabled Control Plane Networks, Anonymous
Defensive Publications Series
The present disclosure proposes a mechanism to improve network efficiency and resource utilization in Retain Home Path (RHP) enabled control plane networks. With optimized resource allocation, networks will be allowed to accommodate a larger number of services without losing RHP capability of existing services. The disclosure provides a way to work with overbooking of RHP efficiently by selecting protect paths which lead to the lowest spectrum utilization in the network for RHP enabled services. Current RHP solutions do not consider spectrum usage. This can lead to usage of extra spectrum, which could have otherwise been saved.
Automatic Detection Of A Non-Effective Prime, Hp Inc
Automatic Detection Of A Non-Effective Prime, Hp Inc
Defensive Publications Series
Some printheads require some servicing routines during the normal operation in order to keep the printheads in good conditions. Typically, the servicing routines consist of a combination of spitting, where the resistors are activated to spit ink, and blow primes, that consist of applying air pressure on the printhead vent to open the printhead regulator and eject some ink. Typically, the result of performing a non‐effective prime is a defect on the Image Quality, since during the servicing routines, the amount of ink ejected during a prime is lower than the required. This situation could occur if there is an …
Widening The Spectrum For Monochrome Led Displays With Occasional Voltage Spikes, Ian Rickard Muldoon
Widening The Spectrum For Monochrome Led Displays With Occasional Voltage Spikes, Ian Rickard Muldoon
Defensive Publications Series
Monochrome LED displays, such as those used for augmented reality (AR) applications, can be unpleasant to view because of the single color used, which is often quite saturated. Approaches to address the issue with the use of phosphors or of LEDs of multiple colors are ineffective or impractical. This disclosure describes techniques to widen the spectrum of colors displayed via monochrome LED displays by occasionally spiking the voltage driving the LED to a higher level. The result is a display appearance that is less saturated and softer.
Global Fraud Detect, Simon Hopley
Global Fraud Detect, Simon Hopley
Defensive Publications Series
The present disclosure relates to a method and a system for aggregating card transactions to detect fraud globally, wherein transactions that take place when the issuer’s accounts are known to be in a sleep state condition or comparable condition. The present disclosure suggests collecting data from an issuer, where there is a failure to detect system-wide fraud. Thereafter, the issuers are enabled to share fraud findings that may include more than simply one card network’s cards. Also, the present disclosure enables cross-card program frauds to be detected and stopped considerably more quickly, consequently reducing the size of a fraud and …
Multicast Operations, Administration, And Management (Oam) Techniques Utilizing Protocol Independent Multicast (Pim) Flooding Mechanisms, Ramakrishnan Chokkanathapuram Sundaram, Stig Venaas, Abhiram Srinivasan
Multicast Operations, Administration, And Management (Oam) Techniques Utilizing Protocol Independent Multicast (Pim) Flooding Mechanisms, Ramakrishnan Chokkanathapuram Sundaram, Stig Venaas, Abhiram Srinivasan
Defensive Publications Series
Multicast networks are often complex and to provide a visualization of traffic flows within a multicast network often involves the full knowledge of a distribution tree for the network. Further, isolating problems within a multicast network can involve tracing of multiple nodes across the distribution tree. Techniques presented provide efficient multicast tree discovery through Protocol Independent Multicast (PIM) flooding mechanisms, which can be further used to facilitate network visualizations and fault isolation within a network.
Real-Time Insights-Guided Beforehand Call Status And Quality Prediction Alerts For To-Be-Called Number, Anupam Mukherjee, Vibhor Jain, Rajarathinam Chidambaram
Real-Time Insights-Guided Beforehand Call Status And Quality Prediction Alerts For To-Be-Called Number, Anupam Mukherjee, Vibhor Jain, Rajarathinam Chidambaram
Defensive Publications Series
Based on a number of factors (including the available bandwidth, network issues, or service provider issues), when a call is made that call may or not be successful. Additionally, the quality of a call may also experience a setback due to those factors. Currently, there is no “beforehand” guidance system in place within a call control ecosystem which can alert a user to the probable relevant issues before a call is made. Techniques are presented herein that address that deficiency. Aspects of the presented techniques encompass a Real-time Intelligent Insights Engine as part of a real-time calling solution (within, for …
Method And System For Selecting The Payment Credentials Based On Rewards During Transaction Processing, Deepak Tomar Mr
Method And System For Selecting The Payment Credentials Based On Rewards During Transaction Processing, Deepak Tomar Mr
Defensive Publications Series
The present invention provides a method and system that can provide best cashback reward by dynamically selecting the payment credentials based on the highest reward cashback percentage offered. In an embodiment, the present invention provides the backdoor updates (105) of the categories and the percentage offered by the issuer on a particular card during a transaction.
Estimation Of Distance Of People From Video Camera Without Dedicated Depth Sensors, Anonymous
Estimation Of Distance Of People From Video Camera Without Dedicated Depth Sensors, Anonymous
Defensive Publications Series
The invention detects using computer vision algorithms, for the video conferencing system user, whether in call or outside call, the user has been sitting or standing at one location for the set duration of time and need to stand-up or take a walk. This experience is currently available through wearable technology where if user is wearing smart watch embedded with appropriate sensors, then the watch can detect such scenarios and then users can stand-up and take a walk. The computer vision-based detection methodology can run from any compute resources, be in the camera compute (CPU, GPU, DSP, specialized AI compute) …
E-Motor Bearing Monitoring Of Electric Vehicles With Uncoupling, Axel Unger
E-Motor Bearing Monitoring Of Electric Vehicles With Uncoupling, Axel Unger
Defensive Publications Series
No abstract provided.
Determination Of The Coefficients Of Friction With Respect To The Road Surface In A Vehicle With An Electric Drive System, Axel Unger
Defensive Publications Series
No abstract provided.
Enforcing Qualified Image And Anti-Rollback Protection On A Device Through Ownership Voucher, Jabir Mohammed, Reda Haddad, Bazil Mohammed Ali, Srihari Raghavan
Enforcing Qualified Image And Anti-Rollback Protection On A Device Through Ownership Voucher, Jabir Mohammed, Reda Haddad, Bazil Mohammed Ali, Srihari Raghavan
Defensive Publications Series
Techniques are presented herein that extend an ownership voucher (OV) to carry software artifact hashes. The presence of such hashes supports a number of functionalities, including enforcing the loading of a specific qualified image to given hardware along with rollback prevention; preventing the installation of other released images through a Universal Serial Bus (USB), an installation workflow, or a network; avoiding pre-staging errors as well as administrative errors on loading images; conveying the known good values of software artifacts for a given release to a customer for a specified release; and aiding operational simplicity and automation using Manufacturer Authorized Signing …
Obtaining User Consent And Indicating Progress During A Sequence Of Automated Actions, Keun Soo Yim, Adam Coimbra
Obtaining User Consent And Indicating Progress During A Sequence Of Automated Actions, Keun Soo Yim, Adam Coimbra
Defensive Publications Series
Automation of actions within a software application can improve user experience, e.g., by enabling repetitive complex sequences to be performed with no user effort. Such actions can include high-level operations, such as actions to be performed via the user interface (UI) of the app. Automation is typically achieved by requesting that a trusted automation service perform each of the high-level actions, including UI input in non-privileged apps running in user space, in the desired sequence on behalf of the user. This disclosure describes techniques to display appropriate UI indicators and get user consent when needed when a sequence of higher-level …
Dynamic Onboarding Process Of Map Data Structure On Hazelcast Cluster, Neeraj Deshmukh, Rohit Kesarwani, Aarushi Gupta, Prabu Kadapenthangal Venkatesan
Dynamic Onboarding Process Of Map Data Structure On Hazelcast Cluster, Neeraj Deshmukh, Rohit Kesarwani, Aarushi Gupta, Prabu Kadapenthangal Venkatesan
Defensive Publications Series
The present invention provides a method to dynamically onboard the map data structure on Hazelcast cluster. To reduce the dependency on multiple teams for onboarding process of a map data structure, the present invention creates a repository for the xml file. The application team can create or modify the xml file for new map onboarding or existing map configuration. Once the changes are made, the application team generates the RPM build for the xml file. The platform team reviews the changes and deploy the RPM build in Hazelcast cluster. The dynamic onboarding of map on Hazelcast cluster starts upon installing …
Enriching Audio Signals In Augmented Reality With Spatial Attributes Of Sound, Corville Allen, Andy Lavery
Enriching Audio Signals In Augmented Reality With Spatial Attributes Of Sound, Corville Allen, Andy Lavery
Defensive Publications Series
This disclosure describes augmented reality (AR) techniques that spatially track audio signals to enable AR interactivity based on audio and on the relationship of the objects that the audio is attached to. An AR object responds interactively based on the properties and the classification of spatially-tracked audio as well as the intent and sentiment inferred from the audio. Spatially separated microphones are leveraged to localize sounds to objects. Alternatively, video object detection is combined with audio processing to localize sounds to objects. AR objects are tagged with spatial audio classification information as they move such that audio information of the …