模型压缩经典文章翻译1：（Network Slimming翻译）Network Slimming-Learning Efficient Convolutional Networks ...

一、番外说明

大家好，我是小P，今天给大家带来深度模型压缩经典文献翻译的第一篇，关于通道剪枝的文章，中英对照。此外，对“目标检测/模型压缩/语义分割”感兴趣的小伙伴，欢迎加入QQ群 813221712 讨论交流，进群请看群公告！
点击链接加入群聊【Object Detection】：https://jq.qq.com/?_wv=1027&k=5kXCXF8
说明：由于CSDN排版的原因，此处就只上传了一部分样例，完整版word文档请百度云盘下载或进群群文件下载，敬请谅解！
广告一下：有对翻译经典文献感兴趣的小伙伴可以私聊我！
百度云链接：链接：https://pan.baidu.com/s/1ACOOVH6CHV1aeNdIO81MWA 提取码：esr7

二、正文翻译

Learning Efficient Convolutional Networks through Network Slimming
Zhuang Liu1∗ Jianguo Li2 Zhiqiang Shen3 Gao Huang4 Shoumeng Yan2 Changshui Zhang1
1CSAI, TNList, Tsinghua University 2Intel Labs China 3Fudan University 4Cornell University {liuzhuangthu, zhiqiangshen0214}@gmail.com, {jianguo.li, shoumeng.yan}@intel.com, [email protected], [email protected]
Abstract
The deployment of deep convolutional neural networks (CNNs) in many Real-World applications is largely hindered by their high computational cost. In this paper, we propose a novel learning scheme for CNNs to simultaneously 1) reduce the model size; 2) decrease the run-time memory footprint; and 3) lower the number of computing operations, without compromising accuracy. This is achieved by enforcing channel-level sparsity in the network in a simple but effective way. Different from many existing approaches, the proposed method directly applies to modern CNN architectures, introduces minimum overhead to the training process, and requires no special software/hardware accelerators for the resulting models. We call our approach network slimming, which takes wide and large networks as input models, but during training insignificant channels are automatically identified and pruned afterwards, yielding thin and compact models with comparable accuracy. We empirically demonstrate the effectiveness of our approach with several state-of-the-art CNN models, including VGGNet, ResNet and DenseNet, on various image classification datasets. For VGGNet, a multi-pass version of network slimming gives a 20× reduction in model size and a 5× reduction in computing operations.
在许多实际应用中部署深度卷积神经网络（CNN）很大程度上受到其计算成本高的限制。在本文中，我们提出了一种新的CNNs学习方案，能同时1）减小模型大小; 2）减少运行时内存占用; 3）在不影响准确率的情况下降低计算操作的数量。这种学习方案是通过在网络中进行通道层次稀疏来实现，简单而有效。与许多现有方法不同，我们所提出的方法直接应用于现代CNN架构，引入训练过程的开销最小，并且所得模型不需要特殊软件/硬件加速器。我们将我们的方法称为网络瘦身（network slimming），此方法以大型网络作为输入模型，但在训练过程中，无关紧要的通道被自动识别和剪枝，从而产生精度相当但薄而紧凑（高效）的模型。在几个最先进的CNN模型（包括VGGNet，ResNet和DenseNet）上，我们使用各种图像分类数据集，凭经验证明了我们方法的有效性。对于VGGNet，网络瘦身后的多通道版本使模型大小减少20倍，计算操作减少5倍。
…
…
…