Zenodo integration

TLDR:
Zenodo integration

v0.0.3. Segmentation models

TLDR:
In this release, we ship segmentation models. In each of their architectures, backbone part is shared so that every segmentation model can be used with different backbone.

Architectures:

• DeepLabV3
• U-Net
• LinkNet
• PSPNet
• FPN

Backbones:

• VGG
• ResNet
• MobileNetV2

v0.0.2. Biodatasets and Classification models

TLDR:
This is a fresh, restructured release package compared to v0.0.1. Here, we ship several classification models and biodatasets in PyTorch friendly format.

Models:

• AlexNet
• GoogLeNet
• MobileNetV2
• MobileNetV3
• ResNet
• ShuffleNetV2
• SqueezeNet
• VGG

Biodatasets:

• ChestXrayDataset
• DSB18Dataset
• HistocancerDataset
• RANZCRDataset
• RetinopathyDataset

v0.0.1. First release

TLDR:
This is the very first release. In this release, we ship various baseline models for classification, segmentation, detection, super-resolution and image translation tasks. As well, basis for model trainers and biodatasets are described here. Architectures are not as clean. Please refer to new releases in the future.

Biodatasets:

• ChestXrayDataset
• DSB18Dataset
• HistocancerDataset
• RANZCRDataset
• RetinopathyDataset

Trainers:

• BaseTrainer
• ConvnetTrainer
• GanTrainer

Models:

• DenseNet
• GoogLeNet
• VGG
• ResNet
• MobileNetV2
• ShuffleNetV2
• ViT
• U-Net
• Attention U-Net
• FasterRCNN
• YOLOv3
• CycleGAN
• SRGAN