给CNN加了个动态记忆模块，让模型学会在新场景中复用重要特征。
Squeeze-and-Remember Block
- 通过压缩和记忆机制，选择性保存训练中的关键特征。
- 在ImageNet上提升0.52%准确率，Cityscapes上提升0.20%分割精度。
- 轻量级设计，适合部署在资源受限的图像处理任务中。
卷积神经网络（CNN）在众多机器学习任务中具有重要作用，其结构包含卷积层、丢弃层和残差层等。然而，现有CNN缺乏类似人脑的记忆机制，难以在新场景中有效利用已有知识。为此，本文提出“压缩与记忆”（Squeeze-and-Remember, SR）块，作为新型网络单元，赋予CNN动态记忆功能。该模块在训练阶段有选择地记忆重要特征，并在推理时自适应地重新引入这些特征，从而增强模型在上下文中的判别能力。在ImageNet和Cityscapes数据集上的实验证明，将SR块嵌入ResNet50后，其在ImageNet上的顶1验证准确率比仅使用Dropout2d提升了0.52%；应用于DeepLab v3时，使Cityscapes数据集的平均交并比（mIoU）提高了0.20%。上述改进均以极小的计算开销实现，表明该模块在图像处理任务中具备显著潜力。
原文摘要 · Abstract (English)
Convolutional Neural Networks (CNNs) are important for many machine learning tasks. They are built with different types of layers: convolutional layers that detect features, dropout layers that help to avoid over-reliance on any single neuron, and residual layers that allow the reuse of features. However, CNNs lack a dynamic feature retention mechanism similar to the human brain's memory, limiting their ability to use learned information in new contexts. To bridge this gap, we introduce the "Squeeze-and-Remember" (SR) block, a novel architectural unit that gives CNNs dynamic memory-like functionalities. The SR block selectively memorizes important features during training, and then adaptively re-applies these features during inference. This improves the network's ability to make contextually informed predictions. Empirical results on ImageNet and Cityscapes datasets demonstrate the SR block's efficacy: integration into ResNet50 improved top-1 validation accuracy on ImageNet by 0.52% over dropout2d alone, and its application in DeepLab v3 increased mean Intersection over Union in Cityscapes by 0.20%. These improvements are achieved with minimal computational overhead. This show the SR block's potential to enhance the capabilities of CNNs in image processing tasks.
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。