liu3xing3long / ms-net

MS-Net in Depth Map Super-Resolution by Deep Multi-Scale Guidance, ECCV 2016

MS-Net (Multi-scale network)

This repository (https://github.com/twhui/MS-Net) is the offical release of MS-Net for our paper Depth Map Super-Resolution by Deep Multi-Scale Guidance in ECCV16. It comes with four trained networks (x2, x4, x8, and x16) one hole-filled RGBD training set, and three hole-filled RGBD testing sets (A, B, and C).

To the best of our knowledge, MS-Net is the FIRST convolution neural network which attempts to upsample depth images.

Another repository for MSG-Net (multi-scale guidance network) is also available.

For more details, please visit my project page.

License and Citation

All code and other materials (including but not limited to the paper, figures, and tables) are provided for research purposes only and without any warranty. Any commercial use requires our consent. When using any parts of the code package or the paper (Depth Map Super-Resolution by Deep Multi-Scale Guidance) in your work, please cite the following paper:

@InProceedings{hui16msgnet,      
 author = {Tak-Wai Hui and Chen Change Loy and and Xiaoou Tang},      
 title  = {Depth Map Super-Resolution by Deep Multi-Scale Guidance},      
 booktitle  = {Proceedings of European Conference on Computer Vision (ECCV)},     
 pages = {353--369},  
 year = {2016},      
 url = {http://mmlab.ie.cuhk.edu.hk/projects/guidance_SR_depth.html}}

Dependency

We train our models using caffe and evaluate the results on Matlab.

Installation and Running

You need to install caffe and remeber to complie matcaffe. You can put the folder MSNet-release in caffe/examples. Finally, you need to get into the the directory of examples/MSNet-release/util, and run MSNet.m.

Training data

Please refer to MSG-Net.

Testing data

Testig set is available at the folder MSNet-release/testing sets.