智能科学与技术学报 ›› 2021, Vol. 3 ›› Issue (3): 359-369.doi: 10.11959/j.issn.2096-6652.202137

• 专刊:目标智能检测与识别 • 上一篇    

特征空间中基于半遗传稀疏表示的图像识别

石林瑞, 黄祎婧, 符进武, 郭心悦, 范自柱   

  1. 华东交通大学理学院,江西 南昌 330013
  • 修回日期:2021-08-17 出版日期:2021-09-01 发布日期:2021-09-01
  • 作者简介:石林瑞(1996− ),女,华东交通大学理学院硕士生,主要研究方向为字典学习、迁移学习、图像处理和模式识别
    黄祎婧(1999− ),女,华东交通大学理学院硕士生,主要研究方向为规则学习
    符进武(1998− ),男,华东交通大学理学院硕士生,主要研究方向为行人重识别
    郭心悦(1999− ),女,华东交通大学理学院硕士生,主要研究方向为目标跟踪
    范自柱(1975− ),男,博士,华东交通大学理学院教授、博士生导师,江西省“井岗学者”特聘教授,江西省“百人远航工程”入选者,华东交通大学天佑人才计划第一层次(杰出人才)入选者,华东交通大学数学学科计算机应用技术方向带头人,主要研究方向为人工智能领域的模式识别与机器学习
  • 基金资助:
    国家自然科学基金资助项目(61991401);江西省自然科学基金资助项目(20192ACBL20010)

Sparse representation for image recognition based on semi-genetic algorithm in feature space

Linrui SHI, Yijing HUANG, Jinwu FU, Xinyue GUO, Zizhu FAN   

  1. School of Science, East China Jiaotong University, Nanchang 330013, China
  • Revised:2021-08-17 Online:2021-09-01 Published:2021-09-01
  • Supported by:
    The National Natural Science Foundation of China(61991401);Jiangxi Provincial Natural Science Foundation(20192ACBL20010)

摘要:

经典的稀疏表示分类(SRC)通常是基于求解L1最小化问题的。SRC在原始输入空间中求解L0范数最小化问题,无法很好地获取数据中的非线性信息。为了解决这一问题,应用非线性映射将原始输入数据映射到一个新的高维特征空间,并提出了一种新的基于L0范数的表示方法。在所提方法中,表示测试样本的字典包含两个部分:第一部分固定在测试样本的近邻;第二部分的训练样本通过半遗传算法(SGA)来选择,利用表示误差确定第二部分的表示字典。在所提方法中,如果训练样本和已确定的测试样本的近邻产生最小表示误差,那么这些训练样本将被 SGA 确定为表示字典的第二部分。在一些常用的人脸数据集和一个手写体数据集上的实验表明,所提方法能够获得更好的分类性能。

关键词: 稀疏表示, 图像识别, 特征空间, L0范数, 遗传算法

Abstract:

The typical sparse representation for classification (SRC) is usually based on L1minimization problem.Conceptually, SRC is essentially an L0norm minimization problem solved in the original input space, which cannot capture well the nonlinear information within the data.In order to address this problem, a nonlinear mapping to map the original input data into a new high dimensional feature space was applied, and a new representation approach based on L0norm was proposed.The representing dictionary used to represent the test sample contains two parts in the proposed approach.The first part is fixed to the neighbors of the test sample.The training samples of the second part is chosen by the variation of genetic algorithm (GA), i.e., the semi GA (SGA) algorithm, which exploits the representation error to determine the second part of the representing dictionary.In the approach, if the training samples combining the determined neighbors of the test sample yield the least representation error, these training samples are determined as the second part of the representing dictionary by SGA.Experiments on several popular face databases and one handwritten digit data set demonstrate that the proposed approach can achieve better classification performance.

Key words: sparse representation, image recognition, feature space, L0 norm, genetic algorithm

中图分类号: 

No Suggested Reading articles found!