# danet-pytorch

**Repository Path**: LynnHg/danet-pytorch

## Basic Information

- **Project Name**: danet-pytorch
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2020-05-19
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# danet-pytorch
[pytorch] DANet: Dual Attention Network for Scene Segmentation

# network：
![image](https://github.com/Andy-zhujunwen/danet-pytorch/blob/master/network.png)

# dataset:
cityspaces

# how to train:
write the dataset path in mypath.py and run 
```
python train.py
```
after trainning,it will save a model: danet.pth
## How to Inference：
#run:
```
python inference.py
```

# inference:
## input:
![image](https://github.com/Andy-zhujunwen/danet-pytorch/blob/master/danet/s1.jpeg)
## output:
![image](https://github.com/Andy-zhujunwen/danet-pytorch/blob/master/danet/testjpg.png)