Practical networks for edge devices adopt shallow depth and small convolutional kernels to save memory and computational cost, which leads to a restricted receptive field. Conventional efficient learning methods focus on lightweight convolution designs, ignoring the role of the receptive field in neural network design. In this paper, we propose the Meta-Pooling framework to make the receptive field learnable for a lightweight network, which consists of parameterized pooling-based operations. Specifically, we introduce a parameterized spatial enhancer, which is composed of pooling operations to provide versatile receptive fields for each layer of a lightweight model. Then, we present a Progressive Meta-Pooling Learning (PMPL) strategy for the parameterized spatial enhancer to acquire a suitable receptive field size. The results on the ImageNet dataset demonstrate that MobileNetV2 using Meta-Pooling achieves top1 accuracy of 74.6\%, which outperforms MobileNetV2 by 2.3\%.
翻译:边缘装置实用网络采用浅深度和小型进化内核,以节省内存和计算成本,从而形成一个有限的可接受场。常规高效学习方法侧重于轻量级变速设计,忽视可接受场在神经网络设计中的作用。在本文中,我们提议元-pooling框架,使可接受场能够为轻量网络学习,该网络由基于参数的集合作业组成。具体地说,我们引入一个参数化空间增强器,由集成作业组成,为轻量模型的每一层提供多功能可接受场。然后,我们为参数化空间增强器提出一个渐进式元-pooling学习(PPPL)战略,以获得合适的可接受场大小。图像网络数据集的结果表明,使用Meta-pooling的移动网络2实现了74.6 ⁇ 的顶端1精度,它比移动网络2+2.3 ⁇ 的移动网络2。