DeepLabv3+ built in TensorFlow
This repo attempts to reproduce Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation (DeepLabv3+) in TensorFlow for semantic image segmentation on the PASCAL VOC dataset and Cityscapes dataset. The implementation is largely based on my DeepLabv3 implementation, which was originally based on DrSleep's DeepLab v2 implemantation and tensorflow models Resnet implementation.
You can install the requirements by running
pip install -r requirements.txt.
This project uses the TFRecord format to consume data in the training and evaluation process. Creating a TFRecord from raw image files is pretty straight forward and will be covered here.
Note: This project includes a script for creating a TFRecord for Cityscapes and Pascal VOC, but not other datasets.
In order to download the Cityscapes dataset, you must first register with their website. After this, make sure to download both
gtFine. You should end up with a folder that will be in the structure
+ cityscapes + leftImg8bit + gtFine
Next, in order to generate training labels for the dataset, clone cityScapesScripts project
git clone https://github.com/mcordts/cityscapesScripts.git cd cityscapesScripts
Then from the root of your cityscapes dataset run
# must have $CITYSCAPES_ROOT defined python cityscapesscripts/preparation/createTrainIdLabelImgs.py
Finally, you can now run the conversion script
create_cityscapes_tf_record.pyprovided in this repository.
Once you have the dataset available, you can create tf records for pascal voc by running the following
bash python create_pascal_tf_record.py --data_dir DATA_DIR \ --image_data_dir IMAGE_DATA_DIR \ --label_data_dir LABEL_DATA_DIR
--pre_trained_model. You also need to convert original data to the TensorFlow TFRecord format. Once you have followed all the steps in dataset preparation and created TFrecord for training and validation data, you can start training model as follow:
bash python train.py --model_dir MODEL_DIR --pre_trained_model PRE_TRAINED_MODELHere,
--pre_trained_modelcontains the pre-trained Resnet model, whereas
--model_dircontains the trained DeepLabv3+ checkpoints. If
--model_dircontains the valid checkpoints, the model is trained from the specified checkpoint in
You can see other options with the following command:
bash python train.py --helpFor inference the trained model with
77.31%mIoU on the Pascal VOC 2012 validation dataset is available here. Download and extract to
The training process can be visualized with Tensor Board as follow:
bash tensorboard --logdir MODEL_DIR
To evaluate how model perform, one can use the following command:
bash python evaluate.py --helpThe current best model build by this implementation achieves
77.31%mIoU on the Pascal VOC 2012 validation dataset.
| Network Backbone | train OS | eval OS | SC | mIOU paper | mIOU repo | |:----------------:|:--------:|:-------:|:---:|:-----------:|:----------:| | Resnet101 | 16 | 16 | | 78.85% | 77.31% |
Here, the above model was trained about 9.5 hours (with Tesla V100 and r1.6) with following parameters:
bash python train.py --train_epochs 43 --batch_size 15 --weight_decay 2e-4 --model_dir models/ba=15,wd=2e-4,max_iter=30k --max_iter 30000
To apply semantic segmentation to your images, one can use the following commands:
bash python inference.py --data_dir DATA_DIR --infer_data_list INFER_DATA_LIST --model_dir MODEL_DIRThe trained model is available here. One can find the detailed explanation of mask such as meaning of color in DrSleep's repo.
Pull requests are welcome. - [x] Implement Decoder - [x] Resnet as Network Backbone - [x] Training on cityscapes - [ ] Xception as Network Backbone - [ ] Implement depthwise separable convolutions - [ ] Make network more GPU memory efficient (i.e. support larger batch size) - [ ] Multi-GPU support - [ ] Channels first support (Apparently large performance boost on GPU) - [ ] Model pretrained on MS-COCO - [ ] Unit test