CNN: Computer Vision: models
Classification Models:
Top-1 Error Top-5 Error Year
AlexNet: 40.96 18.24 2014
VGG-16: 26.78 8.69 2014
ResNet-10: 34.69 14.36 2015
ResNet-18: 28.53 9.82 2015
ResNet-34: 24.84 7.80 2015
ResNet-50: 22.28 6.33 2015
InceptionV3: 21.2 5.60 2015
PreResNet-18: 28.43 9.72 2016
PreResNet-34: 24.89 7.74 2016
PreResnet-50: 22.40 6.47 2016
DenseNet-121: 23.48 7.04 2016
DenseNet-161: 22.86 6.44 2016
PyramidNet-101: 21.98 6.20 2016
ResNeXt-14: 30.32 11.46 2016
ResNeXt-26: 24.14 7.46 2016
Xception: 20.97 5.49 2016
InceptionResNetV2: 19.93 4.90 2016
MobileNet: 26.61 8.9 2017
Segmentation Models:
U-Net: 2015
Attention U-Net: 2018
Detection Models:
R-CNN: 2014
Fast R-CNN: 2015
YOLO v1: 2016
SSD: 2016
YOLO v2: 2017
Mask R-CNN: 2017
YOLO v3: 2018
YOLO v4: 2020
https://github.com/gmalivenko/awesome-computer-vision-models