...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Multimedia, IEEE Transactions on >Sampling and Ontologically Pooling Web Images for Visual Concept Learning
【24h】

Sampling and Ontologically Pooling Web Images for Visual Concept Learning

机译：用于视觉概念学习的Web图像采样和本体池化

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

Sufficient training examples are essential for effective learning of semantic visual concepts. In practice, however, acquiring noise-free training examples has always been expensive. Recently the rapid popularization of social media websites, such as Flickr, has made it possible to collect training exemplars without human assistance. This paper proposes a novel and efficient approach to collect training samples from the noisily tagged Web images for visual concept learning, where we try to maximize two important criteria, relevancy and coverage, of the automatically generated training sets. For the former, a simple method named semantic field is introduced to handle the imprecise and incomplete image tags. Specifically, the relevancy of an image to a target concept is predicted by collectively analyzing the associated tag list of the image using two knowledge sources: WordNet corpus and statistics from Flickr.com. To boost the coverage or diversity of the training sets, we further propose an ontology-based hierarchical pooling method to collect samples not only based on the target concept alone, but also from ontologically neighboring concepts. Extensive experiments on three different datasets (NUS-WIDE, PASCAL VOC, and ImageNet) demonstrate the effectiveness of our proposed approach, producing competitive performance even when comparing with concept classifiers learned using expert- labeled training examples.
机译：足够的培训示例对于有效学习语义视觉概念至关重要。但是，实际上，获取无噪声的训练示例一直很昂贵。最近，诸如Flickr之类的社交媒体网站的迅速普及使得无需人工协助即可收集培训样本。本文提出了一种新颖有效的方法，可以从带有噪点的Web图像中收集训练样本以进行视觉概念学习，在此我们尝试使自动生成的训练集的两个重要标准（相关性和覆盖率）最大化。对于前者，引入了一种简单的名为语义字段的方法来处理不精确和不完整的图像标签。具体而言，通过使用两个知识源（WordNet语料库和Flickr.com的统计信息）共同分析图像的关联标签列表，可以预测图像与目标概念的相关性。为了提高训练集的覆盖面或多样性，我们进一步提出了一种基于本体的分层池化方法，不仅可以仅基于目标概念，还可以从本体上相邻的概念收集样本。在三个不同的数据集（NUS-WIDE，PASCAL VOC和ImageNet）上进行的广泛实验证明了我们提出的方法的有效性，即使与使用专家标签的训练示例所学习的概念分类器进行比较时，也可以产生竞争性能。

著录项

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