对比CNN与ViT在卫星图像分类中的表现，发现数据平衡能缩小模型差距但CNN仍更高效。
Convolutional Neural Nets vs Vision Transformers: A SpaceNet Case Study with Balanced vs Imbalanced Regimes
- 在相同条件下比较EfficientNet-B0与ViT-Base在两类标签分布下的性能。
- 不平衡数据下两模型准确率均达93%，但CNN参数更少、延迟更低；平衡数据下CNN提升至99%。
- 研究揭示数据分布对模型表现影响显著，适合关注模型效率与数据偏移的实践者。
我们在SpaceNet数据集上对卷积神经网络（EfficientNet-B0）与视觉变换器（ViT-Base）进行受控对比，分别在自然不平衡的五类划分和每类700张图像的平衡重采样划分（70:20:10训练/验证/测试）下评估。在匹配预处理（224x224，ImageNet归一化）、轻量级增强和单块NVIDIA P100上的40轮训练预算下，报告了准确率、宏平均F1、平衡准确率、各类别召回率及部署指标（模型大小与延迟）。在不平衡划分中，EfficientNet-B0达到93%测试准确率，具有强宏F1和更低延迟；ViT-Base表现相当（93%），但参数更多、运行时间更长。在平衡划分中，两者均表现优异：EfficientNet-B0达99%，而ViT-Base仍具竞争力，表明平衡可缩小架构差异，但CNN在效率上仍占优。我们公开了清单、日志和逐图像预测结果以支持复现。
原文摘要 · Abstract (English)
We present a controlled comparison of a convolutional neural network (EfficientNet-B0) and a Vision Transformer (ViT-Base) on SpaceNet under two label-distribution regimes: a naturally imbalanced five-class split and a balanced-resampled split with 700 images per class (70:20:10 train/val/test). With matched preprocessing (224x224, ImageNet normalization), lightweight augmentations, and a 40-epoch budget on a single NVIDIA P100, we report accuracy, macro-F1, balanced accuracy, per-class recall, and deployment metrics (model size and latency). On the imbalanced split, EfficientNet-B0 reaches 93% test accuracy with strong macro-F1 and lower latency; ViT-Base is competitive at 93% with a larger parameter count and runtime. On the balanced split, both models are strong; EfficientNet-B0 reaches 99% while ViT-Base remains competitive, indicating that balancing narrows architecture gaps while CNNs retain an efficiency edge. We release manifests, logs, and per-image predictions to support reproducibility.
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。