三维模型 检索 代码_文章导读|一种基于拉普拉斯算子和联合贝叶斯模型的三维形状检索方法...

特征分析在计算机视觉和计算机图形学中起着重要的作用。在形状检索的任务中，形状描述符必不可少。近年来，基于深度学习的特征提取非常流行，但是由于各种形状所包含的内在信息及其可解释性，对它们的几何形状描述符的设计仍然很有意义。

本文提出了一种有效且鲁棒的三维模型描述符，基于表面拉普拉斯-贝尔特拉米算子的归一化特征函数的概率分布和用于降维的频谱方法来构造描述符，利用联合贝叶斯模型来估计描述符空间中的距离，在训练阶段引入矩阵正则化过程来重新估计协方差矩阵。最后，将3D形状检索描述符应用公共基准库。实验表明，该方法鲁棒，具有良好的检索性能。

关键词：

拉普拉斯-贝尔特拉米算子；联合贝叶斯模型；形状检索

Fig.1.  The pipeline of our approach.

Fig.2. Visualization of the IHs of size 32 × 32. The three 3D models undergone isometric transformation.

Fig.3.  The first 32 basis functions in the V-system of degree 3. confidence intervals.

Fig. 4. Retrieval performance of different descriptor under the framework in this paper.

Fig. 5. Results of non-rigid shape retrieval by using VLD, SIHKS and shapeDNA respectively. Each row represents the retrieval result using the 3D model in blue as the query model. Top 8 retrieval results are listed in the figure, and incorrect results are marked with red circles.

Highlights

• A shape descriptor is developed based on the V-transformation of the Laplace-Beltrami eigenfunctions.

• The metric of the descriptor space is learned by the joint Bayesian method and matrix regularization.

• Experiments show the robustness and effectiveness of the developed descriptor and the learned metric.

全文信息

3D shape retrieval based on Laplace operator and joint Bayesian model

BY:  Zihao Wang，Hongwei Lin

Abstract: Feature analysis plays a significant role in computer vision and computer graphics. In the task of shape retrieval, shape descriptor is indispensable. In recent years, feature extraction based on deep learning becomes very popular, but the design of geometric shape descriptor is still meaningful due to the contained intrinsic information and interpretability. This paper proposes an effective and robust descriptor of 3D models. The descriptor is constructed based on the probability distribution of the normalized eigenfunctions of the Laplace-Beltrami operator on the surface, and a spectrum method for dimensionality reduction. The distance metric of the descriptor space is learned by utilizing the joint Bayesian model, and we introduce a matrix regularization in the training stage to re-estimate the covariance matrix. Finally, we apply the descriptor to 3D shape retrieval on a public benchmark. Experiments show that our method is robust and has good retrieval performance.

Keywords: Laplace-Beltrami operator，Joint Bayesian，Shape retrieval

Link: https://doi.org/10.1016/j.visinf.2020.08.002

END

推荐阅读

1、不同任务背景下对配有文本修饰的叙事式可视化的凝视行为分析

2、自动信息图和可视化推荐综述

3、DataV: 面向大尺度高分辨率显示装置的敏捷可视化设计与实现

4、影响因子公布，计算机科学类看这里！

5、TVseer：一个电视收视率的可视分析系统

Visual Informatics

可视信息学

An Open Access Journal

网址：elsevier.com/locate/visinf

邮箱：visinf@cad.zju.edu.cn

长按扫码可关注

点击此处“阅读全文”查看下载原文