DeepLearnLab
DeepLearnLab is an accumulation of Pytorch implementations of milestone papers in Deep Learning. The implementations were developed by Boosung Kim. The site is based off of the moonwalk Jeykll theme. Access through a computer is recommended, as code portions are not mobile-friendly yet.
Image Classification
-
EfficientNet
EfficientNet from "EfficientNet - Rethinking Model Scaling for Convolutional Neural Networks"
-
MobileNetV2
MobileNetV2 from "MobileNetV2 - Inverted Residuals and Linear Bottlenecks"
-
SE-Net
SE-Net from "Squeeze-and-Excitation Networks"
-
DenseNet
DenseNet from "Densely Connected Convolutional Networks"
-
ResNet
ResNet and bottleneck skip-connections from "Deep Residual Learning for Image Recognition"
-
VGG
Various VGG models from "Very Deep Convulutional Networks For Large-Scale Image Recognition"