qwenAILearn

计算机视觉AI技术学习的代码记录

环境:macos12.0.1 + pycharm 2021.1.1 + python3.8 + tensorflow2.6 + annconda

1.各个模型之间的简单性能对比(相同的分类数据，相同的学习率，相同的epoch)

vgg16 14714688 119566341 537.2MB 2762s 0.24 0.42s(预测错了)
AlexNet 3747200 54554629 233.2MB 238s 0.24 0.12s(预测错了)
GoogeNetv1 5880402 / 23.7MB 577s 0.52 0.51s
ResNet50 23571397 / 94.5MB 1091s 0.70 0.52s
MobileNetV1 2419013 / 9.9MB 386s 0.76 0.24s MobileNetV2 2230277 / 9.3MB 442s 0.77 0.33s
MobileNetV3 4208437 / 17.5MB 490s 0.77 0.59s(预测错了)
ShuffleNet2 1258729 / 5.6MB 276s 0.76 0.47s
ShuffleNoFC 2308329 / 9.8MB 221s 0.79 0.47s
DarkNet-19 16872874 / 67.7MB 508s 0.74 0.23s
DarkNet-53 13284970 / 53.3MB 1053s 0.68 0.16s
CSPDarkNet 26678260 / 107MB 太长了 0.74 1.04s
EfficientV1 3996097 / 16.6MB 860s 0.74 0.59(预测错了)
EfficientV2 20183893 / 82MB 1906s 0.22 1.32(预测错了)

Name		Name	Last commit message	Last commit date
Latest commit History 16 Commits
.idea		.idea
AlexNet		AlexNet
DarkNet		DarkNet
EfficientNet		EfficientNet
GoogleNet		GoogleNet
LeNet5		LeNet5
MobileNet		MobileNet
ResNet50		ResNet50
ShuffleNet		ShuffleNet
VGGSSD		VGGSSD
Vgg16		Vgg16
LICENSE		LICENSE
README.md		README.md
class_indices.json		class_indices.json
main.py		main.py
readH5.py		readH5.py
split_train_val.py		split_train_val.py