专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明专利

AU2021101325A4 Methodology for the Detection of Bone Fracture in Humans Using Neural Network 未知
公开(公告)号：AU2021101325A4
公开(公告)日：2021-05-13
申请号：AU2021101325
申请日：2021-03-14
申请人： B LAKSHMI , F FAREEZA , J SEETHA , JAIN RITURAJ , M MADIAJAGAN , M K JAYANTHI KANNAN , SHUKLA ANAND , SHUKLA SAURABH , T SUBRAMANI , THANKAM THAMARI
发明人： F FAREEZA , B LAKSHMI , THANKAM THAMARI , JAIN RITURAJ , SHUKLA ANAND , SHUKLA SAURABH , M K JAYANTHI KANNAN , M MADIAJAGAN , T SUBRAMANI , J SEETHA
IPC分类号： G06T7/00 , A61B5/00 , A61B6/00 , G06K9/46 , G06N3/02 , G16H50/20
摘要： Every day, new and rapidly evolving technologies appear in a number of fields, especially in the medical sector. However, some older strategies are still commonly used, reliable, and useful in this regard. One of these methods for identifying bone fractures is X-rays. However, often the scale of fractures is negligible and cannot be readily identified. As a consequence, powerful and intelligent systems should be developed. The aim of this innovation is to establish an intelligent classification device that can detect and diagnose bone fractures. The evolved method is split into two phases. The images of the fractures are analyzed using various image processing techniques in order to detect their location and shapes in the first stage, and then a back propagation neural network is trained and evaluated on the processed images in the second stage. The method was checked on multiple bone fracture images in the lab, and the findings demonstrated high performance and a high classification rate. Ori inal mage Redced Size3 Layer BackBoeFatr Orgia Img eucdSz Propagation Neural BoeFatr (256X256) image (32X32) Network Detection Resultant Image Fig. 1

2. 发明专利

AU2021100738A4 A Multi-Faceted Optimization Algorithm for Deep Learning Alternative to Gradient Descent 未知
公开(公告)号：AU2021100738A4
公开(公告)日：2021-04-22
申请号：AU2021100738
申请日：2021-02-06
申请人： BABUKARTHIK R G , CHANDRIAH KIRAN KUMAR , GOTHANIA JAISHRI , JAIN RITURAJ , KALEEM , KUMAR PARVENDRA , M MADIAJAGAN , PARIKH SWAPNIL M , PUROHIT NITIN , SHUKLA ANAND , SUBRAMANI T
发明人： BABUKARTHIK R G , KUMAR PARVENDRA , CHANDRIAH KIRAN KUMAR , JAIN RITURAJ , PUROHIT NITIN , KALEEM MOHAMMED KHALID , PARIKH SWAPNIL M , GOTHANIA JAISHRI , SUBRAMANI T , SHUKLA ANAND , M MADIAJAGAN
IPC分类号： G06N3/00 , G06F17/10 , G06N3/02 , G06N7/02 , G06Q10/04
摘要： A Multi-Faceted Optimization Algorithm for Deep Learning Alternative to Gradient Descent Gradient Descent is the popular optimization algorithm used in deep learning and it is stated as a first-order optimization algorithm. The parameters updates are performed in first order derivatives. The direction of the steepest descent is identified by updating the parameters in the opposite direction of the objective function. The major disadvantage is that it will slow the process for the larger dataset and once all the examples processed, then only learning is performed. To overcome the above drawback, as an alternative to gradient descent, the Huddle PSO is proposed. The Huddle PSO is a modified form of PSO, where a single swarm is divided into multiple sub swarms and it is compatible with optimization problems having multiple constraints, to balance both exploration and exploitation. Each sub swarm focuses on a specific region, diversification method is to choose and initiate the sub swarms, the main aim is to increase diversity and concentrate on the optimal value. TSP is solved using PSO by preserving diversity that is by updating the capacity of memory to every particle in PSO and thereby performance is improved for smaller TSP problems. In this invention, huddle PSO is proposed as an alternative to gradient descent comprising the huddling techniques; it states how the main swarm and sub swarm communicates with all other particles, using topological structures. The performance of the proposed algorithm is validated by solving the TSP, and it is compared with the standard TSP dataset taken from standard TSP lib. The performance criteria (computation time, optimal solution, convergence rate, convergence diversity, average convergence) is measured for the proposed techniques. The proposed technique produces better results with a minimum error rate, computation time, and iteration.

3. 发明专利

AU2021103444A4 AIML Based Smart Classifier in a Shared Memory Multiprocessor System 未知
公开(公告)号：AU2021103444A4
公开(公告)日：2022-05-05
申请号：AU2021103444
申请日：2021-06-18
申请人： ARYA PRADEEP , GUPTA SUNIL , SHREE TANU , SINGH ARUN , GUPTA PRATEEK , PATTANAYAK HIMANSU , KUMAR MANOJ , SHUKLA ANAND , SINGH SATYENDR , SHARMA SUNIL KUMAR , SAXENA SANDEEP
发明人： ARYA PRADEEP , GUPTA SUNIL , SHREE TANU , SINGH ARUN , GUPTA PRATEEK , PATTANAYAK HIMANSU , KUMAR MANOJ , SHUKLA ANAND , SINGH SATYENDR , SHARMA SUNIL KUMAR , SAXENA SANDEEP
IPC分类号： G06F9/54 , G06N20/00
摘要： AIML based Smart Classifier in a Shared Memory Multiprocessor System: Separating Data into their Corresponding Class using Biometric Dataset Machine Learning Classification Device. "AIML based Smart Classifier in a Shared Memory Multiprocessor System" is a system that is aimed at shifting the documentation of content through a biometric method and also includes instructions for receiving at least one labeled seed document. This technology receives unlabelled documents with at least one predetermined cost factor training a transductive classifier using the at least one predetermined cost factor, at least one seed document, and unlabelled documents and also classifies the unlabelled documents having a confidence level more than a set limit into several categories using the classifier. Re-categorizing at least some of the existing documents into the categories using the classifier and outputting identifiers to at least one of a user, another system, and another process. The system for separating documents is also shown. Systems and articles of manufacture for searching documents are shown alongside it. A method and system for creating a decision-tree classifier corresponding to a shared-memory multiprocessor system are revealed and the processors primarily create a list for each record attribute in the shared memory. Each attribute list is then allocated to a processor. The processors freely ascertain the best splits for their particular assigned lists, and compliantly find the best global split from all attribute lists. These lists are then allocated again to the processors and divided on the basis of the best global split into the lists for child nodes. Through this invention, the split attribute lists are allocated again to the processors and the process is continuously repeated for each new child node till the point when each attribute list for the new child nodes comprises tuples of a fixed number or from the same record class. This technology also develops new systems, methods, and software that enables an easier manual classification of headnotes and documents and complex headnotes and/or other required data. An excellent system offers a visual UI (user interface) that simultaneously showcases a random headnote of a ranked list of one or more candidate classes along with adjacent classes of the classification system. TOTAL NO OF SHEET: 03 NO OF FIG: 03 FIGI1ET E 1 LSIIATIO FULBLDDT NAREEN IHOEICRAINO H COTATO SN CLDCOST FACTOR: H OTO FO IGA

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式