arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2305.00161cs.CV
收藏

ViewFormer:面向多视图3D形状理解的视图集注意力机制

ViewFormer: View Set Attention for Multi-view 3D Shape Understanding

Renmin University of China(中国人民大学)

Hongyu Sun, Yongcai Wang, Peng Wang, Xudong Cai, Deying Li

AI总结:

本研究提出多视图3D形状理解模型ViewFormer,通过视图集视角与自适应注意力聚合多视图信息,仅用少量参数就在多个数据集和基准任务上刷新了识别与检索性能纪录。

摘要:

本文提出ViewFormer,这是一种用于多视图3D形状识别与检索的简洁且高效的模型。我们系统研究了现有的多视图信息聚合方法,提出了一种新颖的“视图集”视角,该视角最小化了关于视图的关系假设,释放了表示的灵活性。我们设计了一种自适应注意力模型,用于捕捉视图集中元素的成对及高阶相关性。学习到的多视图相关性被聚合成一个具有表达力的视图集描述符,用于识别和检索任务。实验表明,所提出的方法在不同任务和数据集上展现出惊人的性能。例如,仅使用2个注意力块和480万(4.8M)可学习参数,ViewFormer首次在ModelNet40上达到98.8%的识别准确率,比此前的最优方法高出1.1%。在具有挑战性的RGBD数据集上,我们的方法实现了98.4%的识别准确率,相比最强基线有4.1%的绝对提升。ViewFormer还在SHREC'17基准定义的3D形状检索的多个评估维度上创下了新纪录。

英文摘要:

This paper presents ViewFormer, a simple yet effective model for multi-view 3d shape recognition and retrieval. We systematically investigate the existing methods for aggregating multi-view information and propose a novel ``view set" perspective, which minimizes the relation assumption about the views and releases the representation flexibility. We devise an adaptive attention model to capture pairwise and higher-order correlations of the elements in the view set. The learned multi-view correlations are aggregated into an expressive view set descriptor for recognition and retrieval. Experiments show the proposed method unleashes surprising capabilities across different tasks and datasets. For instance, with only 2 attention blocks and 4.8M learnable parameters, ViewFormer reaches 98.8% recognition accuracy on ModelNet40 for the first time, exceeding previous best method by 1.1% . On the challenging RGBD dataset, our method achieves 98.4% recognition accuracy, which is a 4.1% absolute improvement over the strongest baseline. ViewFormer also sets new records in several evaluation dimensions of 3D shape retrieval defined on the SHREC'17 benchmark.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑