Abstract
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\%.
Abstract (translated)
对边缘设备的实用网络采用浅景深和小型卷积核以节省内存和计算成本,导致接收域受限。传统的高效学习方法专注于轻量级卷积设计,忽略了接收域在神经网络设计中的作用。在本文中,我们提出了Meta-Pooling框架,使轻量级网络的接收域可学习,它由参数化的卷积操作组成。具体来说,我们介绍了参数化的空间增强器,它由卷积操作组成,为轻量级模型的每个层提供多功能接收域。然后,我们提出了一种渐进式Meta-Pooling学习(PMPL)策略,以获取适当的接收域大小。在ImageNet数据集上,结果表明,使用Meta-Pooling的MobileNetV2达到top1准确率74.6%,比MobileNetV2高2.3%。
URL
https://arxiv.org/abs/2301.10038