- Deeplab v2/v3 with Resnet and Resnext
- Cityscapes, VOC and ADE-20K dataset
Network and BN implementation are taken from in-placeABN.
Check the run.sh file for an example.
All the parameters are available in argparser.py
The default folder for the logs is logs/<opts.dataset>/<opts.name>.
The log is in the tensorboard format.
.
├── README.md
├── argparser.py
├── data
│ ├── Cityscapes -> PUT here your dataset!
├── dataset -> List of implemented datasets
│ ├── __init__.py
│ ├── ade.py
│ ├── cityscapes.py
│ ├── sampler.py
│ ├── transform.py
│ ├── utils.py
│ └── voc.py
├── metrics -> Class to compute metrics (Acc, mIoU, etc)
│ ├── __init__.py
│ └── stream_metrics.py
├── models -> DeepNN backbones (Resnet, Resnext, WiderResnet)
│ ├── __init__.py
│ ├── densenet.py
│ ├── resnet.py
│ ├── resnext.py
│ ├── util.py
│ └── wider_resnet.py
├── modules -> Modules to build networks (DeepLab and Residual Blocks)
│ ├── __init__.py
│ ├── deeplab.py
│ ├── dense.py
│ ├── misc.py
│ └── residual.py
├── pretrained -> PUT here pretrained models (get them from https://github.com/mapillary/inplace_abn#training-on-imagenet-1k)
├── requirements.txt
├── run.py -> Main file
├── run.sh -> Run example
├── segmentation_module.py -> Helper to create Segmentation Network
├── train.py -> Instance model and perform training/validation
└── utils
├── __init__.py
├── logger.py -> Logger class
├── loss.py -> Custom Losses
├── scheduler.py -> Custom Scheduler
└── utils.py
For any issue, contact me or open an issue here!