一种自适应非线性池化方法，提升图像分类准确率
NIRMAL Pooling: An Adaptive Max Pooling Approach with Non-linear Activation for Enhanced Image Classification
- 结合自适应最大池化与ReLU激活，动态调整特征提取
- 在三个数据集上均优于传统最大池化，最高提升1.62个百分点
- 适合需要增强特征表达的复杂图像识别任务
本文提出NIRMAL Pooling，一种用于卷积神经网络（CNN）的新颖池化层，融合自适应最大池化与非线性激活函数，用于图像分类。该名称代表非线性激活、中间聚合、降维、最大值、自适应和局部化。通过根据目标输出维度动态调整池化参数，并在池化后应用修正线性单元（ReLU）激活，NIRMAL Pooling增强了模型鲁棒性和特征表达能力。我们在三个基准数据集上评估其性能：MNIST数字、MNIST时尚和CIFAR-10。NIRMAL Pooling在MNIST数字上达到99.25%（对比最大池化99.12%），在MNIST时尚上达91.59%（对比91.44%），在CIFAR-10上达70.49%（对比68.87%），表现持续领先，尤其在复杂数据集上优势明显。本工作展示了NIRMAL Pooling在多样化图像识别任务中提升CNN性能的潜力，为传统池化方法提供了一种灵活可靠的替代方案。
原文摘要 · Abstract (English)
This paper presents NIRMAL Pooling, a novel pooling layer for Convolutional Neural Networks (CNNs) that integrates adaptive max pooling with non-linear activation function for image classification tasks. The acronym NIRMAL stands for Non-linear Activation, Intermediate Aggregation, Reduction, Maximum, Adaptive, and Localized. By dynamically adjusting pooling parameters based on desired output dimensions and applying a Rectified Linear Unit (ReLU) activation post-pooling, NIRMAL Pooling improves robustness and feature expressiveness. We evaluated its performance against standard Max Pooling on three benchmark datasets: MNIST Digits, MNIST Fashion, and CIFAR-10. NIRMAL Pooling achieves test accuracies of 99.25% (vs. 99.12% for Max Pooling) on MNIST Digits, 91.59% (vs. 91.44%) on MNIST Fashion, and 70.49% (vs. 68.87%) on CIFAR-10, demonstrating consistent improvements, particularly on complex datasets. This work highlights the potential of NIRMAL Pooling to enhance CNN performance in diverse image recognition tasks, offering a flexible and reliable alternative to traditional pooling methods.
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。