RotateConv: Making Asymmetric Convolutional Kernels Rotatable
文献类型：会议论文
| 作者 | Ma JB(马佳彬)1,2 ; Guo WY(郭韦煜)1; Wang W(王威)1; Wang L(王亮)1
|
| 出版日期 | 2018-08 |
| 会议日期 | August 20-24 2018 |
| 会议地点 | Beijing, China |
| 英文摘要 | In deep Convolutional Neural Networks(CNN), the design of kernel shapes influences a lot on the model size and performance. In this work, our proposed method, RotateConv, applies a novel kernel shape to massively reduce the number of parameters while maintaining considerable performance. The new shape is extremely simple as a line segment one, and we equip it with the rotatable ability which aims to learn diverse features with respect to different angles. The kernel weights and angles are learned simultaneously during end-to-end training via the standard back-propagation algorithm. There are two variants of RotateConv that only have 2 and 4 parameters respectively depending on whether using weight sharing, which are much compressed than the normal 3x3 kernel with 9 parameters. In experiments, we validate our RotateConv with two classical models, ResNet and DenseNet, on four image classification benchmark datasets, namely MNIST, CIFAR10, CIFAR100 and SVHN. |
| 语种 | 英语 |
| URL标识 | 查看原文 |
| 源URL | [http://ir.ia.ac.cn/handle/173211/20885]  |
| 专题 | 自动化研究所_智能感知与计算研究中心 |
| 作者单位 | 1.中国科学院自动化研究所 2.中国科学院大学 |
| 推荐引用方式 GB/T 7714 | Ma JB,Guo WY,Wang W,et al. RotateConv: Making Asymmetric Convolutional Kernels Rotatable[C]. 见:. Beijing, China. August 20-24 2018. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。