Paper

Trinh Man Hoang, Jinjia Zhou, Yibo Fan.

2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Seattle, WA, USA, 2020, pp. 619-623.

The code is built on DSSLIC & PSPNet.

1. Requirements and Dependencies
2. Testing Pre-trained Models
3. Citation

• Ubuntu 16.04.5 LTS
• Python 3.6.8
• Cuda & Cudnn (We test with Cuda = 10.0 and Cudnn = 7.6.5)
• PyTorch 1.3.0
• MATLAB R2018b

Download pretrained models from https://drive.google.com/drive/folders/1lDUPbsYKiBZnCthhKADqwmo2jvop5avz?usp=sharing and put them into the collated folders.

Perform the encoder-decoder matched compression:

$ python test.py --dataroot </path/to/your/imageFolder/> --label_nc 151 --resize_or_crop none --batchSize 1 --gpu_ids 0 --checkpoints_dir checkpoints/ --results_dir </results/path/> --sMapWeights_path ./checkpoints/SMap_epoch_149.pth --fmt png

Perform the residual and down-sampled version compression:

• Download "Binary BPG distribution for Windows (64 bit only)" from https://bellard.org/bpg and put all the binary files in the folder ./evaluation code/bpg-win64

• Download "FLIF Encoder" from https://github.com/FLIF-hub/FLIF and put all the installed binary files in folder ./evaluation code/FLIF-master

• Then perform the residual and down-sampled version compression by using MATLAB and run ./evaluation code/main.m

If you find the code useful in your research, please cite:

@InProceedings{9150905,
    author={T. M. {Hoang} and J. {Zhou} and Y. {Fan}},
    booktitle={2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)},
    title={Image Compression with Encoder-Decoder Matched Semantic Segmentation},
    year={2020},
    volume={},
    number={},
    pages={619-623},
    doi={10.1109/CVPRW50498.2020.00088}
}

Trinh Man Hoang

This repository (as well as its materials) is for non-commercial uses and research purposes only.