/SENet

Reimplementation of image classification model Squeeze-n-Excite Net

Primary LanguagePython

Squeeze and Excitation Networks in TF.keras

Reimplementation of Image Classification model Squeeze and Excitation Networks in Keras 2.0.3+

Note: this was meant to be a practice and not some serious stuff

Models

Current supported models:

  • SE-ResNetV2. Custom ResNets can be built using the SEResNet model builder, whereas prebuilt Resnet models such as SEResNet18, SEResNet50, SEResNet101 and SEResNet154 can also be built directly.