└── README.md


/README.md:
--------------------------------------------------------------------------------
 1 | # A Survey on 3D Deep Learning
 2 | 
 3 | * Dataset
 4 | * Paper
 5 | * Tutorial
 6 | * Packages
 7 | 
 8 | ## Background
 9 | 
10 | #### Data Representation
11 | The 3D data can be represented in the following forms:
12 | * multi-view RGB(D) images
13 | * volumetric
14 | * polygonal mesh
15 | * point cloud
16 | * primitive-based CAD models
17 | 
18 | ## Dataset
19 | 
20 | * [Princeton Segmentation Benchmark dataset (PSB)](http://segeval.cs.princeton.edu/)
21 | 
22 | * [Labeled PSB dataset (L-PSB)](https://people.cs.umass.edu/~kalo/papers/LabelMeshes/)
23 | 
24 | * [Shape COSEG dataset](http://irc.cs.sdu.edu.cn/~yunhai/public_html/ssl/ssd.htm)
25 | 
26 | ## Packages
27 | 
28 | * Geometric Deep Learning Extension Library for PyTorch ([Github](https://github.com/rusty1s/pytorch_geometric))
29 | 
30 | * Pytorch implementation of Graph Convolution Networks & Graph Attention Convolutional Networks ([Github](https://github.com/meliketoy/graph-cnn.pytorch))
31 | 
32 | * PyTorch implementation of Graph Convolutional Networks for semi-supervised classification ([Github](https://github.com/tkipf/pygcn), [paper](https://arxiv.org/abs/1609.02907), [blog](http://tkipf.github.io/graph-convolutional-networks/))
33 | 
34 | * Point cloud semantic segmentation via Deep 3D Convolutional Neural Network ([Github](https://github.com/nsavinov/semantic3dnet), [slides](https://drive.google.com/file/d/0B2SXn94m2RhEbnFMbHVZNVJ3bzA/view))
35 | 
36 | ## Tutorial
37 | 
38 | * A Tutorial on 3D Deep Learning (CVPR 2017) ([web](http://3ddl.stanford.edu/), [video](https://www.youtube.com/watch?v=8CenT_4HWyY))
39 | 
40 | * Machine Learning for 3D Data (Stanford CS468 - Spring 2017) ([web](http://graphics.stanford.edu/courses/cs468-17-spring/schedule.html))
41 | 
42 | * Machine Learning for 3D Data (CSE291-I00 - Winter 2018)([web](https://cse291-i.github.io/schedule.html))
43 | 
44 | * Data-Driven Shape Analysis and Processing ([original version](https://people.cs.umass.edu/~kalo/papers/EGstar16/data_driven_shape.pdf), [latest version](https://people.cs.umass.edu/~kalo/datadrivenshape/data_driven_shape.pdf))
45 | 
46 | * Polygon Mesh Processing ([web](http://www.pmp-book.org/))
47 | 
48 |   A free online book about mesh representation
49 | 
50 | * Udacity Interactive 3D Graphics [web](https://classroom.udacity.com/courses/cs291)
51 | 
52 | ## Papers
53 |   
54 | ### Feature based methods
55 |   
56 | * [2015] 3D Mesh Labeling via Deep Convolutional Neural Networks ([paper](https://dl.acm.org/citation.cfm?id=2835487))
57 |   
58 | * 3D Shape Segmentation via Shape Fully Convolutional Networks ([paper](https://arxiv.org/abs/1702.08675))
59 |   
60 | * 3D Mesh Segmentation via Multi-branch 1D Convolutional Neural Networks ([paper](https://arxiv.org/abs/1705.11050))
61 | 
62 | ### Octree
63 |   
64 | * [2017] O-CNN: Octree-based Convolutional Neural Networks for 3D Shape Analysis ([paper](https://wang-ps.github.io/O-CNN.html), [code](https://github.com/Microsoft/O-CNN))
65 | 
66 | ### Projection
67 | 
68 | * [2017] 3D Shape Segmentation with Projective Convolutional Networks ([paper](https://arxiv.org/abs/1612.02808), [web](https://people.cs.umass.edu/~kalo/papers/shapepfcn/), [code](https://github.com/kalov/ShapePFCN))
69 | 


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