weterence / nndl-semanticsegmentation-cityscape

语义分割-Cityscape

本项目实现了FCN和ENet模型，基于Cityscape论文复现了以下两篇论文：Fully Convolutional Networks for Semantic Segmentation；A Deep Neural Network Architecture for Real-Time Semantic Segmentation

项目小组成员：靳建华 20210980147，张晓琛 20210980070，马嘉晨 20210980109，付涵 20210980124

环境与依赖

python 3.7
kiwisolver>=1.0.1
matplotlib>=3.0.2
numpy>=1.16.0
Pillow>=6.2.0
pyparsing>=2.3.1
python-dateutil>=2.7.5
pytz>=2018.9
six>=1.12.0
pytorch 1.1.1
tensorboardX
tensorborad
cycler>=0.10.0
Image
tqdm
requests

代码结构：

FCN模型的相关代码和预测结果放在"nndl-SemanticSegmentation-Cityscape/FCN/"文件夹下，具体怎么运行FCN相关代码请参考"nndl-SemanticSegmentation-Cityscape/FCN/README.md"

ENet模型的相关代码和预测结果放在"nndl-SemanticSegmentation-Cityscape/ENet/"文件夹下，具体怎么运行FCN相关代码请参考"nndl-SemanticSegmentation-Cityscape/ENet/README.md"

代码使用

1. Train for FCN

python train.py --model fcn32s --backbone vgg16 --dataset pascal_voc --lr 0.0001 --epochs 50

此处train.py文件为"nndl-SemanticSegmentation-Cityscape/FCN/scripts/train.py"

2. Evaluation and test for FCN

python eval.py --model fcn32s --backbone vgg16 --dataset citys

此处train.py文件为"nndl-SemanticSegmentation-Cityscape/FCN/scripts/eval.py"

3. Train for ENet

python main.py -m train --save-dir save/folder/ --name model_name --dataset name --dataset-dir path/root_directory/

此处main.py文件为"nndl-SemanticSegmentation-Cityscape/ENet/main.py"

4. Evaluation and test for ENet

python main.py -m test --save-dir save/folder/ --name model_name --dataset name --dataset-dir path/root_directory/

此处main.py文件为"nndl-SemanticSegmentation-Cityscape/ENet/main.py"

References

code:

• awesome-semantic-segmentation-pytorch
• PyTorch-ENet

paper:

• Fully Convolutional Networks for Semantic Segmentation
• A Deep Neural Network Architecture for Real-Time Semantic Segmentation