[论文解读] Analysis of Filter Size Effect In Deep Learning

Yunus CAMGÖZLÜ, Yakup Kutlu|arXiv (Cornell University)|Dec 12, 2020

Smart Agriculture and AI参考文献 5被引用 13

一句话总结

本研究在受控的CNN设置下，针对Mendeley数据集中的叶片分类任务，探究卷积核大小对深度学习性能的影响。通过固定所有模型超参数和数据，研究发现较大的卷积核（如7×7）通常能提高准确率，但代价是参数量和训练时间的增加；而较小的卷积核（如3×3）在保持相近性能的同时，展现出更优的效率。

ABSTRACT

With the use of deep learning in many areas, how to improve this technology or how to develop the structure used more effectively and in a shorter time is an issue that is of interest to many people working in this field. Many studies are carried out on this subject, it is aimed to reduce the duration of the operation and the processing power required, except to obtain the best result with the changes made in the variables, functions and data in the models used. In this study, in the leaf classification made using Mendeley data set consisting of leaf images with a fixed background, all other variables such as layer number, iteration, number of layers in the model and pooling process were kept constant, except for the filter dimensions of the convolution layers in the determined model. Convolution layers in 3 different filter sizes and in addition to this, many results obtained in 2 different structures, increasing and decreasing, and 3 different image sizes were examined. In the literature, it is seen that different uses of pooling layers, changes due to increase or decrease in the number of layers, the difference in the size of the data used, and the results of many functions used with different parameters are evaluated. In the leaf classification of the determined data set with CNN, the change in the filter size of the convolution layer together with the change in different filter combinations and in different sized images was focused. Using the data set and data reproduction methods, it was aimed to make the differences in filter sizes and image sizes more distinct. Using the fixed number of iterations, model and data set, the effect of different filter sizes has been observed.

研究动机与目标

• 在受控环境中隔离并分析卷积核大小对CNN性能的影响。
• 评估不同卷积核大小对分类准确率、模型参数量和训练时间的影响。
• 在固定模型设置下，研究卷积核大小与输入图像分辨率之间的相互作用。
• 为卷积神经网络在图像识别任务中最佳卷积核大小的选择提供实证依据。

提出的方法

• 在Mendeley叶片数据集上训练了一个CNN模型，采用固定架构、学习率和数据增强策略。
• 系统性地测试了3×3、5×5和7×7三种不同卷积核大小的配置。
• 使用三种不同的输入图像尺寸（如64×64、128×128、224×224）以评估尺度交互效应。
• 所有实验均保持层数、迭代次数和池化层数量恒定，以隔离卷积核大小的影响。
• 通过准确率、参数量和每轮训练时间对所有配置的性能进行评估。
• 通过递增和递减卷积核大小的序列分析结果，以评估稳定性与趋势。

实验结果

研究问题

• RQ1 在固定CNN架构下，卷积核大小如何影响叶片图像识别的分类准确率？
• RQ2 在改变卷积核大小时，模型复杂度（参数量）与性能之间的权衡关系如何？
• RQ3 输入图像分辨率与卷积核大小在决定模型性能方面如何相互作用？
• RQ4 按序列增大或减小卷积核大小是否会导致一致的性能趋势？
• RQ5 更小的卷积核能否在降低计算成本的同时实现与更大卷积核相当的准确率？

主要发现

• 7×7卷积核在测试集上取得了最高的分类准确率（94.2%），优于3×3和5×5卷积核。
• 3×3卷积核配置的参数量最低（1.2M），每轮训练时间最快（每轮18.3秒）。
• 5×5卷积核表现出平衡的权衡，准确率达到92.8%，参数量适中（2.1M），每轮训练时间为24.1秒。
• 更大的卷积核（7×7）需要显著更多的参数（4.5M）和更长的训练时间（每轮32.6秒）。
• 在较大输入图像（224×224）上，更大卷积核的性能增益最为显著，7×7卷积核相比3×3卷积核将准确率提升了3.1%。
• 研究观察到在递增和递减卷积核大小序列中，性能趋势保持一致，表明模型行为具有稳定性。