Rotation-Invariant Convolution Networks with Hexagon-Based Kernels

DOI Web Site 24 References Open Access
  • TANG Yiping
    Faculty of Information Science and Electrical Engineering, Kyushu University
  • HATANO Kohei
    Faculty of Information Science and Electrical Engineering, Kyushu University Riken AIP
  • TAKIMOTO Eiji
    Faculty of Information Science and Electrical Engineering, Kyushu University

Search this article

Description

<p>We introduce the Hexagonal Convolutional Neural Network (HCNN), a modified version of CNN that is robust against rotation. HCNN utilizes a hexagonal kernel and a multi-block structure that enjoys more degrees of rotation information sharing than standard convolution layers. Our structure is easy to use and does not affect the original tissue structure of the network. We achieve the complete rotational invariance on the recognition task of simple pattern images and demonstrate better performance on the recognition task of the rotated MNIST images, synthetic biomarker images and microscopic cell images than past methods, where the robustness to rotation matters.</p>

Journal

References(24)*help

See more

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top