Research on Image Information Mining Needed for Multi-modal Knowledge Graph Construction and Application-Reference-Cited by-同舟云学术

Research on Image Information Mining Needed for Multi-modal Knowledge Graph Construction and Application

Published:2024-01-30 Issue: Volume: Page:
ISSN:
Container-title:
language:
Short-container-title:

Author:

Peng Jinghui¹,Hu Xinyu²,Yang Jian³,Li Yi¹

Affiliation:

1. Anhui Polytechnic University

2. Sun Yat-sen University

3. Xidian University

Abstract

Abstract An image is an important form of information transmission and contains a lot of effective information. With the explosive growth of multi-modal data represented by pictures, the multi-modal knowledge graph (MMKG) has become an effective means to manage and apply. It is necessary to obtain comprehensive and effective image data to construct a high-quality MMKG. This research focuses on the construction of MMKG, mainly from the analysis of graph structure and characteristics. Firstly, the structural characteristics and elements composition of the MMKG are described. Then, introduced the existing forms of image entity recognition, multi-features capture, scene graphs generation respectively, and description text generation is in the graph and summarized the main mining methods. Finally, analyzed several applications of image data in a commodity multi-modal knowledge graph.

Publisher

Research Square Platform LLC

Reference44 articles.

1. Image-Driven Data Mining for Image Content Segmentation, Classification, and Attribution[J];Barnes C;IEEE Trans Geosci Remote Sensing,2009

2. OmarBonergePinedaLezama, Patio-Saucedo J, NoelVarela, et al. Convolutional Neural Network with Multi-column Characteristics Extraction for Image Classification[J]. Cham: Springer; 2020.

3. Zhang X, Xiong H, Zhou W et al. Picking Deep Filter Responses for Fine-Grained Image Recognition[C]// Computer Vision & Pattern Recognit IEEE, 2016:1134–42.

4. Yi S, Wang X, Tang X. Deep Learning Face Representation from Predicting 10,000 Classes[C]// IEEE Conference on Computer Vision & Pattern Recognition. IEEE, 2014.

5. Katsurai M, Ogawa T, Haseyama M. A cross-modal approach for extracting semantic relationships of concepts from an image database[C]// IEEE International Conference on Acoustics. IEEE, 2012.