Multimodal Learning Using Haar Scattering Transform

WANG Juan, ZHAO Jie

doi:10.13371/j.cnki.chin.q.j.m.2019.03.009

Chinese Quarterly Journal of Mathematics >

2019 , Vol. 34 >Issue 3: 314 - 322

DOI: https://doi.org/10.13371/j.cnki.chin.q.j.m.2019.03.009

Multimodal Learning Using Haar Scattering Transform

Expand

College of Science, Zhongyuan University of Technology

WANG Juan(1984-), female, native of Shangqiu, Henan, a lecturer of Zhongyuan University of Technology, Ph.D., engages in image processing; ZHAO Jie(1983-), male, native of Kaifeng, Henan, a lecturer of Zhongyuan University of Technology, Ph.D., engages in partial of differential equations.

Received date: 2019-06-10

Online published: 2020-08-23

Supported by

Supported by the Natural Science Foundation of China(11626239)； China Scholarship Council(No.201708410483)； the Foundation of Education Department of Henan Province(18A110037)；

Fold

Abstract

In this paper, we propose a novel application of Haar scattering transform to learn features over multiple modalities data. A series of tasks for multimodal learning are presented, and the way of multimodal feature learning is shown. Furthermore, we validate our methods on several datasets with an classification task, demonstrating that the approach is effective.

Key words： Multimodal learning; Haar scattering transform; Classification task

Cite this article

WANG Juan, ZHAO Jie . Multimodal Learning Using Haar Scattering Transform[J]. Chinese Quarterly Journal of Mathematics, 2019 , 34(3) : 314 -322 . DOI: 10.13371/j.cnki.chin.q.j.m.2019.03.009

Options

Outlines

模态框（Modal）标题

Abstract

Cite this article