I am a little nervous about the Dataset and Dataloader. I really want to see that how you organize the dataset.Can you release the ModelNet40 dataset of different modalities, including image, point-cloud and mesh modalities. I am looking forward to your reply.

piseabhijeet commented 3 years ago

Hi,

You can use modelnet dataset from:

Meshnet repo for meshes
Images using the original multi view moelnet dataset
for point clouds, you can save the vertices of the meshes.

LongLong-Jing commented 3 years ago

@huacong @piseabhijeet Sorry for the late reply.

Thanks for your interest in our work. You can download all the data that used in our paper from this link: https://www.dropbox.com/sh/2um7io1g69w7elf/AABix9QyR63UFN3-20HHNXEAa?dl=0.

There are five files including: modelnet10_hdf5_2048.zip contains the point cloud data for ModelNet10 dataset. Link: https://www.dropbox.com/s/3hx4egrfayxekmi/modelnet10_hdf5_2048.zip?dl=0

ModelNet10-Images-180.tar contains the images for ModelNet10 dataset. Link: https://www.dropbox.com/s/beds8f6fpawymkn/ModelNet10-Images-180.tar?dl=0

ModelNet40_Mesh.tar contains the mesh data for ModelNet40 dataset. Link: https://www.dropbox.com/s/dkqhbxhbxz4v9uw/ModelNet40_Mesh.tar?dl=0

modelnet40_ply_hdf5_2048.zip contains the point cloud data for ModelNet40 dataset. Link: https://www.dropbox.com/s/t5swwtd8ecqr7j5/modelnet40_ply_hdf5_2048.zip?dl=0

ModelNet40-Images-180.zip contains the image for Modelnet40 dataset. Link: https://www.dropbox.com/s/byhr0a5jmkg1cpp/ModelNet40-Images-180.zip?dl=0

duzhenjiang113 commented 3 years ago

Hi, i meet some problems about the dataset of the ModelNet40, that is :

ile "/home/anaconda3/envs/pqnet/lib/python3.6/site-packages/torch/utils/data/_utils/worker.py", line 178, in _worker_loop data = fetcher.fetch(index) File "/home/anaconda3/envs/pqnet/lib/python3.6/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/anaconda3/envs/pqnet/lib/python3.6/site-packages/torch/utils/data/_utils/fetch.py", line 44, in data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/Cross-Modal-Center-Loss-main/tools/triplet_dataloader.py", line 219, in getitem centers, corners, normals, neighbor_index = self.get_mesh(item) File "/home/Cross-Modal-Center-Loss-main/tools/triplet_dataloader.py", line 168, in get_mesh data = np.load(mesh_path) File "/homeanaconda3/envs/pqnet/lib/python3.6/site-packages/numpy/lib/npyio.py", line 444, in load pickle_kwargs=pickle_kwargs) File "/home/anaconda3/envs/pqnet/lib/python3.6/site-packages/numpy/lib/npyio.py", line 193, in init _zip = zipfile_factory(fid) File "/home/anaconda3/envs/pqnet/lib/python3.6/site-packages/numpy/lib/npyio.py", line 119, in zipfile_factory return zipfile.ZipFile(file, *args, **kwargs) File "/homeanaconda3/envs/pqnet/lib/python3.6/zipfile.py", line 1108, in init self._RealGetContents() File "/home/anaconda3/envs/pqnet/lib/python3.6/zipfile.py", line 1175, in _RealGetContents raise BadZipFile("File is not a zip file") zipfile.BadZipFile: File is not a zip file

it seems there are some problems about ModelNet40-mesh

duzhenjiang113 commented 3 years ago

Could you please tell me how to transform the mesh from .off to .npz?

LongLong-Jing commented 3 years ago

Hi @duzhenjiang113, You can directly download the mesh data that I uploaded in the Dropbox folder.

You can also generate the zip file from off file based on the code from this repo: https://github.com/iMoonLab/MeshNet

duzhenjiang113 commented 3 years ago

Hi @duzhenjiang113, You can directly download the mesh data that I uploaded in the Dropbox folder.

You can also generate the zip file from off file based on the code from this repo: https://github.com/iMoonLab/MeshNet

Thanks a lot. I find the problem is that the file of tv_stand_0164.nzp is damaged.

huacong commented 3 years ago

@huacong @piseabhijeet Sorry for the late reply.

Thanks for your interest in our work. You can download all the data that used in our paper from this link: https://www.dropbox.com/sh/2um7io1g69w7elf/AABix9QyR63UFN3-20HHNXEAa?dl=0.

There are five files including: modelnet10_hdf5_2048.zip contains the point cloud data for ModelNet10 dataset. Link: https://www.dropbox.com/s/3hx4egrfayxekmi/modelnet10_hdf5_2048.zip?dl=0

ModelNet10-Images-180.tar contains the images for ModelNet10 dataset. Link: https://www.dropbox.com/s/beds8f6fpawymkn/ModelNet10-Images-180.tar?dl=0

ModelNet40_Mesh.tar contains the mesh data for ModelNet40 dataset. Link: https://www.dropbox.com/s/dkqhbxhbxz4v9uw/ModelNet40_Mesh.tar?dl=0

modelnet40_ply_hdf5_2048.zip contains the point cloud data for ModelNet40 dataset. Link: https://www.dropbox.com/s/t5swwtd8ecqr7j5/modelnet40_ply_hdf5_2048.zip?dl=0

ModelNet40-Images-180.zip contains the image for Modelnet40 dataset. Link: https://www.dropbox.com/s/byhr0a5jmkg1cpp/ModelNet40-Images-180.zip?dl=0

Oh, maybe without my rights. I can't download the ModelNet40-Images-180.zip

LongLong-Jing commented 3 years ago

Hi @huacong, thanks for your interest in our work. I checked the link for ModelNet40-Images-180.zip again and I am able to download it. Compared to other files, this file is much larger which may cause the issue. Can you try it again?

LongLong-Jing / Cross-Modal-Center-Loss

I am a little nervous about the Dataset and Dataloader. I really want to see that how you organize the dataset.Can you release the ModelNet40 dataset of different modalities, including image, point-cloud and mesh modalities. I am looking forward to your reply. #5