vinits5/pc_autoencoder

Auto Encoder for 3D Point Clouds using ModelNet40 Dataset

Auto Encoder for 3D Point Clouds

Network Structure:

Code:

Steps to train the auto-encoder:

1. Download ModelNet40 Dataset [Link]
2. Clone repository.
3. Extract the zip file and copy modelnet40_ply_hdf5_2048 folder to pc_autoencoder/data.
4. python pointnet_autoencoder_train.py --mode train

Steps to test the auto-encoder:

1. Download dataset as given in training steps.
2. Download weights for the trained network. [Link]
3. python pointnet_autoencoder_train.py --mode test

Visualise the Dataset: python show_pc.py idx idx: Index of Point Cloud in ModelNet40 Dataset.

Results:

Red colored point clouds are input to the network and blue point clouds are the output.

[Note: A translation has been applied to blue point clouds during testing for a better visualisation purpose.]

Additional Results:

References:

1. PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation [Link]
2. PCN: Point Completion Network [Link]

pc_autoencoder's Issues

model_best.t7

hello
i tried to run the train code but it showed:

wang@deep-learn:~/3ddensenet.torch/examples$ ./run_modelnet40_h5.sh 0,1
nGPU:2
batchSize:48
log is saved into logs/20200820162653_3ddensenet_modelnet_bz48_de30_gr16_modelnet40_60x_h5_n2.log
20200820162656
modelnet
/home/wang/torch/install/bin/luajit: /home/wang/3ddensenet.torch/models/init.lua:30: File not found: /home/wang/3ddensenet.torch/data/data/modelnet40_60x_h5/3ddensenet_modelnet_bz48_de30_gr16/model_best.t7
stack traceback:
[C]: in function 'assert'
/home/wang/3ddensenet.torch/models/init.lua:30: in function 'setup'
main.lua:35: in main chunk
[C]: in function 'dofile'
...wang/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
[C]: at 0x00406510
log is saved into logs/20200820162653_3ddensenet_modelnet_bz48_de30_gr16_modelnet40_60x_h5_n2.log

Where can I get this “model_best.t7 “？
thanks

'module' object has no attribute 'lazy'

hello
i tried to run the train code but it showed:

wang@deep-learn:~/pc_autoencoder$ python pointnet_autoencoder_train.py --mode train
Traceback (most recent call last):
File "pointnet_autoencoder_train.py", line 5, in
import tensorflow as tf
File "/usr/local/lib/python2.7/dist-packages/tensorflow/init.py", line 30, in
from tensorflow._api.v2 import compat
File "/usr/local/lib/python2.7/dist-packages/tensorflow/_api/v2/compat/init.py", line 22, in
from tensorflow._api.v2.compat import v2
File "/usr/local/lib/python2.7/dist-packages/tensorflow/_api/v2/compat/v2/init.py", line 302, in
"Limited tf.compat.v2.summary API due to missing TensorBoard "
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/tools/component_api_helper.py", line 56, in package_hook
child_pkg = importlib.import_module(child_package_str)
File "/usr/lib/python2.7/importlib/init.py", line 37, in import_module
import(name)
File "/usr/local/lib/python2.7/dist-packages/tensorboard/summary/init.py", line 32, in
from tensorboard.summary import v2
File "/usr/local/lib/python2.7/dist-packages/tensorboard/summary/v2.py", line 24, in
from tensorboard.plugins.audio.summary_v2 import audio
File "/usr/local/lib/python2.7/dist-packages/tensorboard/plugins/audio/summary_v2.py", line 30, in
from tensorboard.compat import tf2 as tf
File "/usr/local/lib/python2.7/dist-packages/tensorboard/compat/init.py", line 28, in
import tensorboard.lazy as _lazy
AttributeError: 'module' object has no attribute 'lazy'

How can I solve these problems

cuda version problem

Excuse me, is the environment of this project based on cuda8+?

tf_nndistance_so.so

hello
i tried to run the train code but it showed:

python pointnet_autoencoder_train.py
Traceback (most recent call last):
File "pointnet_autoencoder_train.py", line 79, in
MODEL = importlib.import_module(FLAGS.model) # import network module
File "/usr/lib/python2.7/importlib/init.py", line 37, in import_module
import(name)
File "/home/sun/pc_autoencoder/models/pointnet_pose.py", line 12, in
import tf_util_loss
File "/home/sun/pc_autoencoder/utils/tf_util_loss.py", line 2, in
from pc_distance import tf_nndistance, tf_approxmatch
File "/home/sun/pc_autoencoder/utils/pc_distance/tf_nndistance.py", line 5, in
nn_distance_module=tf.load_op_library(os.path.join(BASE_DIR, 'tf_nndistance_so.so'))
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/load_library.py", line 64, in load_op_library
None, None, error_msg, error_code)
tensorflow.python.framework.errors_impl.NotFoundError: /home/sun/pc_autoencoder/utils/pc_distance/tf_nndistance_so.so: cannot open shared object file: No such file or directory

and i tried to use "makefile" in "pc_distance",it showed:

make: 放弃循环依赖 tf_nndistance.cu <- tf_nndistance.cu.o 。
g++ tf_nndistance.cpp tf_nndistance.cu.o -o tf_nndistance_so.so
-I /usr/local/cuda-8.0/include/ -I /usr/local/lib/python2.7/dist-packages/tensorflow/include -I /usr/local/lib/python2.7/dist-packages/external/nsync/public
-L /usr/local/cuda-8.0/lib64/ -lcudart
-L /usr/local/lib/python2.7/dist-packages/tensorflow/core/lib -ltensorflowframe
-shared -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++11 -fPIC -O2
/usr/bin/ld: 找不到 -ltensorflow
collect2: error: ld returned 1 exit status
makefile:15: recipe for target 'tf_nndistance_so.so' failed
make: *** [tf_nndistance_so.so] Error 1

is it because our environment is different?or something else