dmlc--dgl
Tong He
558673e139
* commit patch * commit patch * pointnet basic * fix data * reorg * reorg * temp status * remove validate set * add partseg data and model * partseg miou * clean up * fix loss * network definition match paper * fix * fix miou * update data format * fix * fix * working pointnet ssg cls * avoid some pytorch bug * fix script * update hyperparams * add msg module * try different dataset * update new dataset info * quick fix to subgraph * fix speed * update training * update * fix bs * update docstring * update * update * remove parallel reduction in fps * switch to kernel fps, training is 30% faster Co-authored-by: Ubuntu <ubuntu@ip-172-31-20-181.us-west-2.compute.internal>
26 行
667 B
Markdown
26 行
667 B
Markdown
Dynamic EdgeConv
|
|
====
|
|
|
|
This is a reproduction of the paper [Dynamic Graph CNN for Learning on Point
|
|
Clouds](https://arxiv.org/pdf/1801.07829.pdf).
|
|
|
|
The reproduced experiment is the 40-class classification on the ModelNet40
|
|
dataset. The sampled point clouds are identical to that of
|
|
[PointNet](https://github.com/charlesq34/pointnet).
|
|
|
|
To train and test the model, simply run
|
|
|
|
```python
|
|
python main.py
|
|
```
|
|
|
|
The model currently takes 3 minutes to train an epoch on Tesla V100, and an
|
|
additional 17 seconds to run a validation and 20 seconds to run a test.
|
|
|
|
The best validation performance is 93.5% with a test performance of 91.8%.
|
|
|
|
## Dependencies
|
|
|
|
* `h5py`
|
|
* `tqdm`
|