基于形变卷积神经网络的手写体数字识别研究
|
摘要
本文引入形变卷积模块来增强网络对数字几何变换的建模能力,提出了一种基于改进的形变卷积神经网络手写体数字识别框架,在提高识别精度的同时,还有效的减少了训练的参数量,提高识别速度.本文在手写体数据集及变换后的数据集中进行验证.实验结果的分析以及与相应算法的比较,证明了本算法是有效的.
In this paper, the deformable convolution module is introduced to enhance the modeling ability of the network to digital geometric transformation, and an improved handwritten digital recognition framework based on deformable CNN is proposed. In addition to improving the recognition accuracy, the framework can effectively reduce the training parameters and improve the recognition speed. This paper demonstrate state-of-the-art performance competing methods on the handwritten dataset and the transformed dataset.
In this paper, the deformable convolution module is introduced to enhance the modeling ability of the network to digital geometric transformation, and an improved handwritten digital recognition framework based on deformable CNN is proposed. In addition to improving the recognition accuracy, the framework can effectively reduce the training parameters and improve the recognition speed. This paper demonstrate state-of-the-art performance competing methods on the handwritten dataset and the transformed dataset.
引文
[1] 陈岩, 李洋洋, 余乐,等. 基于卷积神经网络的手写体数字识别系统[J]. 微电子学与计算机, 2018,35(2):71-74.
[2] KAREN SIONYAN, ANDREW ZISSERMAN. Very deep convolutional networks for large-scale image recognition[J]. Computer Science, 2014, arXiv preprint arXiv:1409.1556.
[3] 赵朋成, 冯玉田, 涂云轩. 基于高倍特征深度残差网络的手写数字识别[J]. 电子测量技术, 2018,6(41):86-89.
[4] YANG F, JIN L, YANG W, et al. Handwritten/Printed Receipt Classification Using Attention-Based Convolutional Neural Network[C]// International Conference on Frontiers in Handwriting Recognition. IEEE, 2017:384-389.
[5] DAI JIFENG, et al. Deformable Convolutional Networks [C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017:764-773.
[6] DENG J, DONG W, Socher R, et al. ImageNet: A large-scale hierarchical image database[C]//IEEE Conference on Computer Vision and Pattern Recognition, 2009:248-255.
[2] KAREN SIONYAN, ANDREW ZISSERMAN. Very deep convolutional networks for large-scale image recognition[J]. Computer Science, 2014, arXiv preprint arXiv:1409.1556.
[3] 赵朋成, 冯玉田, 涂云轩. 基于高倍特征深度残差网络的手写数字识别[J]. 电子测量技术, 2018,6(41):86-89.
[4] YANG F, JIN L, YANG W, et al. Handwritten/Printed Receipt Classification Using Attention-Based Convolutional Neural Network[C]// International Conference on Frontiers in Handwriting Recognition. IEEE, 2017:384-389.
[5] DAI JIFENG, et al. Deformable Convolutional Networks [C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017:764-773.
[6] DENG J, DONG W, Socher R, et al. ImageNet: A large-scale hierarchical image database[C]//IEEE Conference on Computer Vision and Pattern Recognition, 2009:248-255.