对比9种轻量CNN在资源约束下的表现，发现新模型并非全面更优。
Do Newer Lightweight CNNs Perform Better Under Resource Constraints? A Controlled Multigenerational Study of Architecture, Initialization, Training Budget, and Efficiency
- 统一协议下对比9个轻量CNN在多个数据集的表现
- EfficientNet-B0在参数和算力少80%时仍接近最佳性能
- 新模型优势有限，硬件环境显著影响推理速度
新式轻量级卷积神经网络常被宣传为提升预测性能与部署效率，但此类声明需经受严格评估。本研究在共享下游协议下，对九个轻量级CNN模型包在CIFAR-10、CIFAR-100和Tiny ImageNet上的表现进行对比。报告了顶1准确率、宏F1、顶5准确率、参数量、FP32存储、GMACs、NVIDIA L4与AMD Ryzen 5 5500U CPU上批量大小为1的延迟、峰值PyTorch CUDA张量内存占用，以及点估计的帕累托前沿。EfficientNetV2-S在CIFAR-10和CIFAR-100上分别达到最高顶1准确率97.57%和86.98%，RepViT-M1.0在Tiny ImageNet上达到79.87%。EfficientNet-B0在三数据集上分别落后最优结果0.22、0.85、1.79个百分点，但参数量减少约79%，GMACs减少86%，且出现在所有评估的准确率与资源帕累托前沿上，是最具一致竞争力的中等预算选项。MobileNetV3-Small GMACs最低，两种CPU设置下均最快，且在三个数据集上表现优于MobileNetV4-Conv-S。随机初始化下，其领先幅度分别为2.55、1.76、0.99点，固定训练模型的置信区间不包含零。100轮从头训练后，EfficientNet-B0仍比预训练版本低3.29、10.10、17.54点，尽管训练时间约为其五倍。SqueezeNet1.1参数最少，峰值CUDA占用最低，但准确率显著偏低。延迟排序在L4与CPU间差异显著，表明GMACs无法单独预测实际推理性能。总体而言，新设计仅带来选择性而非普遍性提升。
原文摘要 · Abstract (English)
Newer lightweight convolutional neural networks are often presented as improving predictive performance and deployment efficiency, but such claims require controlled evaluation. This study compares nine lightweight CNN model packages across CIFAR-10, CIFAR-100, and Tiny ImageNet under a shared downstream protocol. We report top-1 accuracy, macro F1, top-5 accuracy, parameter count, FP32 storage, GMACs, batch-size-1 latency on an NVIDIA L4 and AMD Ryzen 5 5500U CPU, peak PyTorch CUDA allocated tensor memory, and point estimate Pareto frontiers. EfficientNetV2-S achieves the highest observed top-1 accuracy on CIFAR-10 and CIFAR-100 at 97.57% and 86.98%, while RepViT-M1.0 leads Tiny ImageNet at 79.87%. EfficientNet-B0 remains within 0.22, 0.85, and 1.79 percentage points of the best result on the three datasets while using approximately 79% fewer parameters and 86% fewer GMACs than EfficientNetV2-S. It also appears on every evaluated accuracy and resource Pareto frontier, making it the most consistently competitive intermediate-budget option. MobileNetV3-Small has the lowest GMAC count, is the fastest model under both CPU thread settings, and records higher observed accuracy than MobileNetV4-Conv-S on all three datasets. Under random initialization, it leads MobileNetV4-Conv-S by 2.55, 1.76, and 0.99 points, with paired test-set intervals excluding zero for the fixed trained models. EfficientNet-B0 remains 3.29, 10.10, and 17.54 points below its pretrained counterpart after 100 epochs of scratch training, despite requiring about five times the recorded training time. SqueezeNet1.1 has the fewest parameters and lowest peak CUDA allocation, but substantially weaker accuracy. Latency rankings differ sharply between the L4 and CPU environments, showing that GMACs alone do not predict measured inference performance. Overall, newer designs provide selective rather than universal gains
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。