https://github.com/pochih/FCN-pytorch

🚘 Easiest Fully Convolutional Networks
https://github.com/pochih/FCN-pytorch

computer-vision deep-learning fully-convolutional-networks pytorch semantic-segmentation

Last synced: 2 months ago
JSON representation

🚘 Easiest Fully Convolutional Networks

• Host: GitHub
• URL: https://github.com/pochih/FCN-pytorch
• Owner: pochih
• Created: 2017-10-12T08:42:10.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2023-07-06T21:15:35.000Z (over 1 year ago)
• Last Synced: 2024-08-04T00:02:50.973Z (6 months ago)
• Topics: computer-vision, deep-learning, fully-convolutional-networks, pytorch, semantic-segmentation
• Language: Python
• Homepage:
• Size: 59.1 MB
• Stars: 401
• Watchers: 11
• Forks: 143
• Open Issues: 28
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-semantic-segmentation - pytorch

README

        
[![Open Source Love](https://badges.frapsoft.com/os/v1/open-source-150x25.png?v=103)](https://github.com/ellerbrock/open-source-badges/)

## 🚘 The easiest implementation of fully convolutional networks

- Task: __semantic segmentation__, it's a very important task for automated driving

- The model is based on CVPR '15 best paper honorable mentioned [Fully Convolutional Networks for Semantic Segmentation](https://arxiv.org/abs/1411.4038)

## Results

### Trials



### Training Procedures



## Performance

I train with two popular benchmark dataset: CamVid and Cityscapes

|dataset|n_class|pixel accuracy|

|---|---|---

|Cityscapes|20|96%

|CamVid|32|93%

## Training

### Install packages

```bash

pip3 install -r requirements.txt

```

and download pytorch 0.2.0 from [pytorch.org](pytorch.org)

and download [CamVid](http://mi.eng.cam.ac.uk/research/projects/VideoRec/CamVid/) dataset (recommended) or [Cityscapes](https://www.cityscapes-dataset.com/) dataset

### Run the code

- default dataset is CamVid

create a directory named "CamVid", and put data into it, then run python codes:

```python

python3 python/CamVid_utils.py 

python3 python/train.py CamVid

```

- or train with CityScapes

create a directory named "CityScapes", and put data into it, then run python codes:

```python

python3 python/CityScapes_utils.py 

python3 python/train.py CityScapes

```

## Author

Po-Chih Huang / [@pochih](https://pochih.github.io/)