首页|Multi-label Image Classification via Coarse-to-Fine Attention

Multi-label Image Classification via Coarse-to-Fine Attention

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
  • 维普
Great efforts have been made by using deep neural networks to recognize multi-label images.Since multi-label image classification is very complicated,many studies seek to use the attention mechanism as a kind of guidance.Conventional attention-based methods always analyzed images directly and aggressively,which is difficult to well understand complicated scenes.We propose a global/local attention method that can recognize a multi-label image from coarse to fine by mimicking how human-beings observe images.Our global/local attention method first concentrates on the whole image,and then focuses on its local specific objects.We also propose a joint max-margin objective function,which enforces that the minimum score of positive labels should be larger than the maximum score of negative labels horizontally and vertically.This function further improve our multi-label image classification method.We evaluate the effectiveness of our method on two popular multi-label image datasets (i.e.,Pascal VOC and MS-COCO).Our experimental results show that our method outperforms state-of-the-art methods.

Multi-label classificationConvolutional neural networkRecurrent neural networkAttention

LYU Fan、LI Linyan、Victor S.Sheng、FU Qiming、HU Fuyuan

展开 >

School of Electronic and Information Engineering, Suzhou University of Science and Technology, Suzhou 215009, China

College of Intelligence and Computing, Tianjin University, Tianjin 300072, China

Suzhou Institute of Trade and Commerce, Suzhou 215009, China

Department of Computer Science, Texas Tech University, Lubbock TX 79409, USA

展开 >

This work is supported by the National Natural Science Foundation of ChinaThis work is supported by the National Natural Science Foundation of ChinaThis work is supported by the National Natural Science Foundation of ChinaThis work is supported by the National Natural Science Foundation of ChinaThis work is supported by the National Natural Science Foundation of ChinaPrimary Research & Developement Plan of Jiangsu ProvinceNatural Science Foundation of the Higher Education Institutions of Jiangsu ProvinceFoundation of Key Laboratory in Science and Technology Development Project of SuzhouFoundation of Key Laboratory in Science and Technology Development Project of Suzhou

61876121No.61472267No.61728205No.61502329No.61672371BE201766319KJB520054SZS201609No.SZS201813

2019

10.1049/cje.2019.07.015

中国电子杂志(英文版)

中国电子杂志(英文版)

CSTPCDCSCDSCIEI
ISSN：1022-4653
年,卷(期)：2019.28(6)
  • 55