ColorPolarNet: Residual Dense Network-Based Chromatic Intensity-Polarization Imaging in Low-Light Environment
This repository contains the codes for paper ColorPolarNet: Residual Dense Network Based Chromatic Intensity-Polarization Imaging in Low-Light Environment by Xiuyu Xu, Minjie Wan*, Junyu Ge, et al. (*Corresponding author).
The overall repository style is partially borrowed from IPLNet (https://github.com/Dismoon/IPLNet). Thanks to Haofeng Hu.
We provide the LLCP dataset of our paper. You can download them by the following link. [LLCP (Google Drive)] [LLCP (Baidu Drive pw:85eg)]
- Python 3.6
- Tensorflow 1.13
- CUDA10.0 and cuDNN
- h5py 2.8
- opencv
- Download and release LLCP.zip. Put the LLCP folder in the ColorPolarNet folder.
- Setup the required parameters in the main scripts.
- Run main.py for training or testing.