提出轻量级卷积网络CoSNet，兼顾低计算量与高效率。
Designing Concise ConvNets with Columnar Stages
- 采用并行小核卷积+列式堆叠结构，减少参数和计算量
- 在ImageNet上达到78.6%准确率，仅需2.4G FLOPs
- 适合移动端等资源受限场景部署
在视觉Transformer兴起的背景下，VanillaNet展示了简单紧凑卷积神经网络的巨大潜力。然而，这类模型多聚焦运行时性能，还需同时优化FLOPs、参数量等指标以增强实用性。为此，我们提出一种新型卷积网络宏观设计——列式阶段网络（Columnar Stage Network, CoSNet）。CoSNet具有系统化设计的简洁结构，较浅深度、低参数量、低FLOPs，且无需注意力机制，非常适合资源受限场景部署。其核心创新在于：通过输入复制驱动的并行小核卷积，采用列式堆叠方式整合特征，并最小化1x1卷积层使用。全面评估表明，CoSNet在资源受限条件下可媲美多种知名卷积网络与Transformer架构。代码已开源。
原文摘要 · Abstract (English)
In the era of vision Transformers, the recent success of VanillaNet shows the huge potential of simple and concise convolutional neural networks (ConvNets). Where such models mainly focus on runtime, it is also crucial to simultaneously focus on other aspects, e.g., FLOPs, parameters, etc, to strengthen their utility further. To this end, we introduce a refreshing ConvNet macro design called Columnar Stage Network (CoSNet). CoSNet has a systematically developed simple and concise structure, smaller depth, low parameter count, low FLOPs, and attention-less operations, well suited for resource-constrained deployment. The key novelty of CoSNet is deploying parallel convolutions with fewer kernels fed by input replication, using columnar stacking of these convolutions, and minimizing the use of 1x1 convolution layers. Our comprehensive evaluations show that CoSNet rivals many renowned ConvNets and Transformer designs under resource-constrained scenarios. Code: https://github.com/ashishkumar822/CoSNet
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。