人工智能培训

搜索

人工智能论文:所需要的只是几个转变:为图像分类设计高效的卷积神经网络(All You Need is a Few Shifts: Designing Effic

[复制链接]
bangat 发表于 2019-3-15 13:09:15 | 显示全部楼层 |阅读模式
bangat 2019-3-15 13:09:15 158 0 显示全部楼层
人工智能论文:所需要的只是几个转变:为图像分类设计高效的卷积神经网络(All You Need is a Few Shifts: Designing Efficient Convolutional Neural  Networks for Image Classification)移位操作是深度可分离卷积的有效替代方案。然而,它的实现方式仍然是瓶颈,即记忆运动。为了推动这一方向的发展,本文引入了一种名为稀疏移位层(SSL)的新型基本组件,用于构建高效的卷积神经网络。在这个体系结构中,基本块仅由1x1卷积层组成,只有少量移位操作应用于中间特征映射。为了使这个想法可行,我们在优化过程中引入了移位操作惩罚,并进一步提出了一种量化感知移位学习方法,以使得学习位移对于推理更加友好。广泛的消融研究表明,只有少数班次操作足以提供空间信息通信。此外,为了最大限度地发挥SSL的作用,设计了一种改进的网络架构,以充分利用神经网络(FE-Net)的有限容量。配备SSL,该网络在ImageNet上可实现75.0%的前1精度,仅有563M M-Adds。它超越了由深度可分卷积构造的其他部件和NAS在精度和实际速度方面所研究的网络。
Shift operation is an efficient alternative over depthwise separableconvolution.However, it is still bottlenecked by its implementation manner,namely memory movement.To put this direction forward, a new and novel basiccomponent named Sparse Shift Layer (SSL) is introduced in this paper toconstruct efficient convolutional neural networks.In this family ofarchitectures, the basic block is only composed by 1x1 convolutional layerswith only a few shift operations applied to the intermediate feature maps.Tomake this idea feasible, we introduce shift operation penalty duringoptimization and further propose a quantization-aware shift learning method toimpose the learned displacement more friendly for inference.Extensive ablationstudies indicate that only a few shift operations are sufficient to providespatial information communication.Furthermore, to maximize the role of SSL, weredesign an improved network architecture to Fully Exploit the limited capacityof neural Network (FE-Net).Equipped with SSL, this network can achieve 75.0%top-1 accuracy on ImageNet with only 563M M-Adds.It surpasses othercounterparts constructed by depthwise separable convolution and the networkssearched by NAS in terms of accuracy and practical speed.人工智能论文:所需要的只是几个转变:为图像分类设计高效的卷积神经网络(All You Need is a Few Shifts: Designing Efficient Convolutional Neural  Networks for Image Classification) JDcg0DdmmAagtGgc.jpg
URL地址:https://arxiv.org/abs/1903.05285     ----pdf下载地址:https://arxiv.org/pdf/1903.05285    ----人工智能论文:所需要的只是几个转变:为图像分类设计高效的卷积神经网络(All You Need is a Few Shifts: Designing Efficient Convolutional Neural  Networks for Image Classification)
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则 返回列表 发新帖

bangat当前离线
新手上路

查看:158 | 回复:0

快速回复 返回顶部 返回列表