├── README.md
└── doc
    └── image
        ├── example.png
        ├── logo.gif
        └── title.jpg


/README.md:
--------------------------------------------------------------------------------
 1 | <div align="center">
 2 |   <img src="doc/image/logo.gif" width="300"/>
 3 | </div>
 4 | 
 5 | <div align="center">- A Photography Portrait Matting Benchmark - </div>
 6 | 
 7 | 
 8 | 
 9 | <p align="center">
10 |   <a href="#news">News</a> |
11 |   <a href="#introduction">Introduction</a> |
12 |   <a href="#download">Download</a> | 
13 |   <a href="#license">License</a> | 
14 |   <a href="#citation">Citation</a> | 
15 |   <a href="#contact">Contact</a>
16 | </p>
17 | 
18 | <img src="doc/image/title.jpg" width="100%">
19 | 
20 | ---
21 | 
22 | ## News
23 | - **[Jul 20 2021] PPM-100 Benchmark is Released!**  
24 |   The benchmark with 100 finly-annotated, high-resolution images (PPM-100) is released. 
25 |   
26 | 
27 | ## Introduction
28 | PPM is a portrait matting benchmark with the following characteristics:
29 | - **Fine Annotation** - All images are labeled and checked carefully.
30 | - **Natural Background** - All images use the original background without replacement.
31 | - **Rich Diversity** - The images cover full/half body and various postures.
32 | - **High Resolution** - The resolution of images is between 1080p and 4k.
33 | 
34 | Below is an example image:
35 | 
36 | <img src="doc/image/example.png" width="100%">
37 | 
38 | 
39 | 
40 | 
41 | ## Download
42 | Currently, PPM-100 used in the [MODNet](https://github.com/ZHKKKe/MODNet) paper is available.  
43 | Note that few images used in the MODNet paper are replaced by similar images due to license issues.   
44 | You can download PPM-100 from:  
45 | [Google Drive](https://drive.google.com/file/d/1JUx-EPoV9QAhQgmW0AyOen-xKQUzZia-/view?usp=sharing) | [百度网盘 (提取码: PPMB)](https://pan.baidu.com/s/1iIeowdj6sdKSqW6qGnSb6A)
46 | 
47 | 
48 | ## License
49 | All original portrait images in PPM are from [Flickr](https://www.flickr.com/) and constrained by [Flickr Creative Commons License (Commercial use & mods allowed)](https://www.flickr.com/creativecommons/).  
50 | All annotated alpha mattes in PPM are released under the [Creative Commons Attribution NonCommercial ShareAlike 4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/legalcode) license.  
51 | 
52 | 
53 | ## Citation
54 | If you use this PPM benckmark, please cite:
55 | 
56 | ```bibtex
57 | @InProceedings{MODNet,
58 |   author = {Zhanghan Ke and Jiayu Sun and Kaican Li and Qiong Yan and Rynson W.H. Lau},
59 |   title = {MODNet: Real-Time Trimap-Free Portrait Matting via Objective Decomposition},
60 |   booktitle = {AAAI},
61 |   year = {2022},
62 | }
63 | ```
64 | 
65 | 
66 | ## Contact
67 | This repository is currently maintained by Zhanghan Ke ([@ZHKKKe](https://github.com/ZHKKKe)).  
68 | If there is any question, please contact `kezhanghan@outlook.com`.
69 | 


--------------------------------------------------------------------------------
/doc/image/example.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/ZHKKKe/PPM/459d0bfc456047679aa98972bed1bb5f379cba54/doc/image/example.png


--------------------------------------------------------------------------------
/doc/image/logo.gif:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/ZHKKKe/PPM/459d0bfc456047679aa98972bed1bb5f379cba54/doc/image/logo.gif


--------------------------------------------------------------------------------
/doc/image/title.jpg:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/ZHKKKe/PPM/459d0bfc456047679aa98972bed1bb5f379cba54/doc/image/title.jpg


--------------------------------------------------------------------------------