Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Geometric question answering method and model based on deep learning and multi-modal numerical reasoning

A deep learning and multi-modal technology, applied in inference methods, neural learning methods, electrical and digital data processing, etc., can solve problems such as increased processing time, large amount of processed data, and wrong screening, etc., to increase the accuracy of answers, Improve the efficiency of processing and the effect of good accuracy

Pending Publication Date: 2021-11-19
SUN YAT SEN UNIV SHENZHEN +1
View PDF2 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, the current commonly used method has the following technical problems: due to the many topics involved, and each topic only needs to change the parameters or data, the answer method will change, and more answers will be derived. If only by identifying the image to search for the answer, Answers input by a single user can only be selected from a large number of answers, which is not conducive to students' extended learning, and the amount of data to be processed is large, which increases processing time and reduces processing efficiency. Moreover, if the questions are similar, it is easy to misselect. Reduce the accuracy of screening and affect the user experience

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Geometric question answering method and model based on deep learning and multi-modal numerical reasoning
  • Geometric question answering method and model based on deep learning and multi-modal numerical reasoning
  • Geometric question answering method and model based on deep learning and multi-modal numerical reasoning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0054] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0055]The current commonly used methods have the following technical problems: due to the large number of topics involved, and each topic only needs to change the parameters or data, the answer method will change, and more answers will be derived. Being able to screen a large number of answers to obtain a single user-input answer is not conducive to students' extended learning, and the amount of data to be processed is large, which increases processing time and redu...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a deep learning and multi-modal numerical reasoning-based geometric question answering method and a text and image bimodal combined neural network model. The method comprises the following steps: respectively acquiring text information and image information about question contents; encoding the text information into a corresponding text hiding state to obtain text semantics, and encoding the image information into a corresponding image hiding state to obtain visual semantics; fusing and aligning the text semantics and the visual semantics to obtain an answering program; and calculating an answering result according to the operation mode of the answering program. According to the method, the answering accuracy can be increased, the processing efficiency can be improved, and the technology which has good accuracy and high practicability and can autonomously generate the code sequence of the answer through deep learning is achieved.

Description

technical field [0001] The invention relates to the technical field of intelligent education, in particular to a method for solving geometric problems based on deep learning and multimodal numerical reasoning, and a neural network model combining text and image dual modes. Background technique [0002] With the development and popularization of artificial intelligence, artificial intelligence has been applied to various industries, one of which is intelligent education. [0003] At present, one of the most commonly used applications is intelligent answering. Its operation method is that the user takes a picture of the corresponding question, and by identifying the content of the question in the picture, based on the content of the question, he searches in the large question bank built from a large number of questions to find the corresponding answer. Answer. [0004] However, the current commonly used method has the following technical problems: due to the many topics invol...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/332G06F16/33G06F40/30G06N3/04G06N3/08G06N5/04
CPCG06F16/3329G06F16/3344G06F40/30G06N3/08G06N5/04G06N3/044G06N3/045
Inventor 梁小丹李橦李奇文陈嘉奇
Owner SUN YAT SEN UNIV SHENZHEN
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products