arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2108.12952cs.CV
收藏

通过极匹配卷积学习发现反射对称性

Learning to Discover Reflection Symmetry via Polar Matching Convolution

Pohang University of Science and Technology (POSTECH)(浦项科技大学)

Ahyun Seo, Woohyeon Shim, Minsu Cho

摘要:

反射对称检测的任务仍然具有挑战性,由于显着的变化和模糊的对称模式在野外。此外,由于局部区域需要在反射中匹配以检测对称模式,因此标准卷积网络很难学习该任务,因为标准卷积网络与旋转和反射不是等变的。为了解决这个问题,我们引入了一种新的卷积技术,称为极坐标匹配卷积,它利用了极坐标特征池,自相似编码和不同角度轴的系统内核设计。提出的高维核卷积网络有效地学习从真实世界的图像中发现对称模式,克服了标准卷积的局限性。此外,我们提出了一个新的数据集,并引入了一个自监督学习策略,通过合成图像来增强数据集。实验表明,我们的方法优于国家的最先进的方法在准确性和鲁棒性。

英文摘要:

The task of reflection symmetry detection remains challenging due to significant variations and ambiguities of symmetry patterns in the wild. Furthermore, since the local regions are required to match in reflection for detecting a symmetry pattern, it is hard for standard convolutional networks, which are not equivariant to rotation and reflection, to learn the task. To address the issue, we introduce a new convolutional technique, dubbed the polar matching convolution, which leverages a polar feature pooling, a self-similarity encoding, and a systematic kernel design for axes of different angles. The proposed high-dimensional kernel convolution network effectively learns to discover symmetry patterns from real-world images, overcoming the limitations of standard convolution. In addition, we present a new dataset and introduce a self-supervised learning strategy by augmenting the dataset with synthesizing images. Experiments demonstrate that our method outperforms state-of-the-art methods in terms of accuracy and robustness.

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。

↑