Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/cardwing/codes-for-intra-kd
Inter-Region Affinity Distillation for Road Marking Segmentation (CVPR 2020)
https://github.com/cardwing/codes-for-intra-kd
cnn deep-learning lane-detection pytorch road-marking-segmentation
Last synced: 16 days ago
JSON representation
Inter-Region Affinity Distillation for Road Marking Segmentation (CVPR 2020)
- Host: GitHub
- URL: https://github.com/cardwing/codes-for-intra-kd
- Owner: cardwing
- License: mit
- Created: 2020-03-02T02:35:14.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2020-04-14T07:19:32.000Z (over 4 years ago)
- Last Synced: 2023-10-20T18:40:39.066Z (about 1 year ago)
- Topics: cnn, deep-learning, lane-detection, pytorch, road-marking-segmentation
- Language: Python
- Homepage:
- Size: 781 KB
- Stars: 108
- Watchers: 5
- Forks: 21
- Open Issues: 16
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
Codes for ["Inter-Region Affinity Distillation for Road Marking Segmentation"](https://arxiv.org/abs/2004.05304)
## Requirements
- [PyTorch 0.3.0](https://pytorch.org/get-started/previous-versions/)
- Opencv
- cvbase
## Before start
Please follow [list](./list) to put ApolloScape in the desired folder. We'll call the directory that you cloned Codes-for-IntRA-KD as `$IntRA_KD_ROOT .
## Testing
1. Obtain model predictions from trained weights:
Download the trained [ResNet-101](https://drive.google.com/open?id=16TJW4K69uSb_ChBlbqX33aJ7LP43dhKf) and [ERFNet](https://drive.google.com/open?id=145B-xNl89R7H9qEZ6r8TzK-KSG6jf0Dp), and put them in the folder ```trained_model```.
```
cd $IntRA_KD_ROOT
sh test_pspnet_multi_scale.sh # sh test_erfnet_multi_scale.sh
```
The output predictions will be saved to ```road05_tmp``` by default.
2. Transfer TrainID to ID:
```
python road_npy2img.py
```
The outputs will be stored in ```road05``` by default.
3. Generate zip files:
```
mkdir test
mv road05 test/
zip -r test.zip test
```
Now, just upload test.zip to [ApolloScape online server](http://apolloscape.auto/submit.html). The trained ResNet-101 can achieve **46.63%** mIoU and trained ERFNet can achieve **43.48%** mIoU.
4. (Optional) Produce color maps from model predictions:
```
python trainId2color.py
```
5. (Optional) Leverage t-SNE to visualize the feature maps:
Please use the [script](https://github.com/cardwing/Codes-for-Steering-Control/blob/master/tools/draw_tsne.py) to perform the visualization.
## Training
```
cd $IntRA_KD_ROOT
sh train_pspnet.sh # sh train_erfnet_vanilla.sh
```
Please make sure that you have 8 GPUs and each GPU has least 11 GB memory if you want to train ResNet-101.
## Citation
If you use the codes, please cite the following publication:
```
@inproceedings{hou2020interregion,
title = {Inter-Region Affinity Distillation for Road Marking Segmentation},
author = {Yuenan Hou, Zheng Ma, Chunxiao Liu, Tak-Wai Hui, and Chen Change Loy},
booktitle = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2020},
}
```
## Acknowledgement
This repo is built upon [ERFNet-CULane-PyTorch](https://github.com/cardwing/Codes-for-Lane-Detection).
## Contact
If you have any problems in reproducing the results, just raise an issue in this repo.
## To-Do List
- [ ] Training codes of IntRA-KD and various baseline KD methods for ApolloScape