CoConv: Learning Dynamic Cooperative Convolution for Image Recognition
CoConv: Learning Dynamic Cooperative Convolution for Image Recognition
复制标题
DOI：
10.1109/icme51207.2021.9428105
复制
发表时间：
2021-07
期刊：
影响因子：
--
通讯作者：
Kien X. Nguyen;Tiffany Ryu;Jocelyn Zhang;Xu Ma;Qing Yang;Song Fu;P. Palacharla;N. Wang;Xi Wang
中科院分区：
文献类型：
--
作者：
Kien X. Nguyen;Tiffany Ryu;Jocelyn Zhang;Xu Ma;Qing Yang;Song Fu;P. Palacharla;N. Wang;Xi Wang
在本文中，我们提出了一个概念简单，但功能强大的图像识别方法。该方法被称为协作动态卷积（CoConv），它引入了来自多个卷积专家的动态卷积的协作学习。CoConv可以作为传统静态卷积的替代品，并且可以无缝集成到各种视觉模型中。此外，CoConv很容易训练，在推理阶段只引入了最小的计算开销。CoConv通过同时使用多个卷积专家进行训练，并且在卷积操作之前通过加权求和合并卷积权重，以提高推理过程中的效率。大量实验的结果表明，CoConv在各种数据集上的图像分类方面都有一致的改进，与基础卷积网络的选择无关。值得注意的是，CoConv在ImageNet上将ResNet18的前1分类准确率提高了3.06%。该代码可从以下网址获得：https://github.com/Nyquixt/CoConv。
In this paper, we present a conceptually simple, yet powerful method for image recognition. The method, called Cooperative Dynamic Convolution (CoConv), introduces a cooperative learning of dynamic convolution from multiple convolutional experts. CoConv can be used as a substitute for the traditional static convolution, and can be seamlessly integrated in various visual models. Moreover, CoConv is easy to train with only a minimal computational overhead introduced in the inference phase. CoConv is trained by using multiple convolutional experts simultaneously, and the convolutional weights are merged by a weighted summation before convolutional operations for efficiency during inference. Results from extensive experiments show that CoConv leads to consistent improvement for image classification on various datasets, independent of the choice of the base convolutional network. Remarkably, CoConv improves the top-1 classification accuracy of ResNet18 by 3.06% on ImageNet. The code is available at: https://github.com/Nyquixt/CoConv.