• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Computer vision and image understanding >Enhanced representation and multi-task learning for image annotation
【24h】

Enhanced representation and multi-task learning for image annotation

机译：图像表示的增强表示和多任务学习

获取原文
获取原文并翻译 | 示例
       
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In this paper we propose a novel biased random sampling strategy for image representation in Bag-of-Words models. We evaluate its impact on the feature properties and the ranking quality for a set of semantic concepts and show that it improves performance of classifiers in image annotation tasks and increases the correlation between kernels and labels. As second contribution we propose a method called Output Kernel Multi-Task Learning (MTL) to improve ranking performance by transfer information between classes. The main advantages of output kernel MTL are that it permits asymmetric information transfer between tasks and scales to training sets of several thousand images. We give a theoretical interpretation of the method and show that the learned contributions of source tasks to target tasks are semantically consistent. Both strategies are evaluated on the ImageCLEF PhotoAnnotation dataset. Our best visual result which used the MTL method was ranked first according to mean Average Precision (mAP) within the purely visual submissions in the ImageCLEF 2011 PhotoAnnotation Challenge. Our multi-modal submission achieved the first rank by mAP among all submissions in the same competition.
机译：在本文中，我们提出了一种新颖的有偏随机抽样策略，用于单词袋模型中的图像表示。我们评估了它对一组语义概念的特征属性和排名质量的影响，并表明它提高了图像标注任务中分类器的性能，并增加了内核和标签之间的相关性。作为第二贡献，我们提出了一种称为输出内核多任务学习（MTL）的方法，以通过在类之间传递信息来提高排名性能。输出内核MTL的主要优点在于，它允许任务之间的不对称信息传输，并缩放到数千个图像的训练集。我们对该方法进行了理论解释，并表明源任务对目标任务的学习贡献在语义上是一致的。两种策略都在ImageCLEF PhotoAnnotation数据集上进行评估。在ImageCLEF 2011 PhotoAnnotation挑战的纯视觉提交中，使用MTL方法获得的最佳视觉效果在平均视觉精度（mAP）中排名第一。我们的多模式提交方式在同一比赛的所有提交方式中均获得了mAP的第一名。

著录项

  • 来源
    《Computer vision and image understanding》 |2013年第5期|466-478|共13页
  • 作者

    Alexander Binder; Wojciech Samek; Klaus-Robert Mueller; Motoaki Kawanabe;

  • 作者单位

    Fraunhofer Institute FIRST, Kekulestr. 7, 12489 Berlin, Germany Machine Learning Croup, Berlin Institute of Technology (TU Berlin), Marchstrasse 23, 10587 Berlin, Germany;

    Fraunhofer Institute FIRST, Kekulestr. 7, 12489 Berlin, Germany Machine Learning Croup, Berlin Institute of Technology (TU Berlin), Marchstrasse 23, 10587 Berlin, Germany;

    Machine Learning Croup, Berlin Institute of Technology (TU Berlin), Marchstrasse 23, 10587 Berlin, Germany Bernstein Focus: Neurotechnology Berlin, 10587 Berlin, Germany Department of Brain and Cognitive Engineering, Korea University, Anam-dong, Seongbuk-gu, Seoul 136-713, Republic of Korea;

    ATR Brain Information Communication Research Laboratory Croup, 2-2-2 Hikaridai, Seika-cho, Soraku-gun,Kyoto 619-0288, Japan Fraunhofer Institute FIRST, Kekulestr. 7, 12489 Berlin, Germany;

  • 收录信息
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

    Image ranking; Image classification; Multiple kernel learning; Multi task learning; Bag-of-Words representation; Biased random sampling; ImageCLEF; Mutual information;

    机译：图片排名;图像分类;多核学习;多任务学习;词袋表示法;偏向随机抽样;ImageCLEF;相互信息;
  • 入库时间 2022-08-17 13:21:07

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号