人工智能培训

搜索

人工智能论文:用于非刚性三维形状检索的多特征距离度量学习(Multi-feature Distance Metric Learning for Non-rigi

[复制链接]
1449231467 发表于 2019-1-11 10:29:38 | 显示全部楼层 |阅读模式
1449231467 2019-1-11 10:29:38 261 0 显示全部楼层
人工智能论文:用于非刚性三维形状检索的多特征距离度量学习(Multi-feature Distance Metric Learning for Non-rigid 3D Shape Retrieval)在过去的几十年中,基于特征学习的三维形状检索方法在计算机图形社区中得到了广泛的关注。这些方法通常探索手工制作的距离度量或常规距离度量学习方法来计算单个特征的相似性。单个特征总是包含一个几何信息,不能很好地表征3D形状。因此,应该将多个特征用于检索任务,以克服单个特征的限制并进一步提高性能。然而,大多数常规距离度量学习方法未能将来自多个特征的互补信息集成以构建距离度量。为了解决这些问题,本研究提出了一种新的多特征距离度量学习方法,用于非刚性三维形状检索,利用KL-散度可以充分利用多个形状特征的互补几何信息。在不同的特征度量和公共度量之间最小化KL-散度是一致性约束,这可以导致多特征的一致性共享潜在特征空间。我们将该方法应用于三维模型检索,并在众所周知的基准数据库上测试我们的方法。结果表明,我们的方法基本上优于最先进的非刚性3D形状检索方法。
In the past decades, feature-learning-based 3D shape retrieval approacheshave been received widespread attention in the computer graphic community.These approaches usually explored the hand-crafted distance metric orconventional distance metric learning methods to compute the similarity of thesingle feature.The single feature always contains onefold geometricinformation, which cannot characterize the 3D shapes well.Therefore, themultiple features should be used for the retrieval task to overcome thelimitation of single feature and further improve the performance.However, mostconventional distance metric learning methods fail to integrate thecomplementary information from multiple features to construct the distancemetric.To address these issue, a novel multi-feature distance metric learningmethod for non-rigid 3D shape retrieval is presented in this study, which canmake full use of the complimentary geometric information from multiple shapefeatures by utilizing the KL-divergences.Minimizing KL-divergence betweendifferent metric of features and a common metric is a consistency constraints,which can lead the consistency shared latent feature space of the multiplefeatures.We apply the proposed method to 3D model retrieval, and test ourmethod on well known benchmark database.The results show that our methodsubstantially outperforms the state-of-the-art non-rigid 3D shape retrievalmethods.人工智能论文:用于非刚性三维形状检索的多特征距离度量学习(Multi-feature Distance Metric Learning for Non-rigid 3D Shape Retrieval) n6brMRGlBBaJ1J1w.jpg
URL地址:https://arxiv.org/abs/1901.03031     ----pdf下载地址:https://arxiv.org/pdf/1901.03031    ----人工智能论文:用于非刚性三维形状检索的多特征距离度量学习(Multi-feature Distance Metric Learning for Non-rigid 3D Shape Retrieval)
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则 返回列表 发新帖

1449231467当前离线
新手上路

查看:261 | 回复:0

快速回复 返回顶部 返回列表