Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Wrong results with my own data #5

Open
ProgrammingMonkey123 opened this issue Mar 2, 2021 · 0 comments
Open

Wrong results with my own data #5

ProgrammingMonkey123 opened this issue Mar 2, 2021 · 0 comments

Comments

@ProgrammingMonkey123
Copy link

ProgrammingMonkey123 commented Mar 2, 2021

I am trying to upsample lossy compressed point cloud data with PU-Net and I get four separate different-scales objects.
The data is 8i-people dataset with MPEG TMC13 lossy compressed. Pity it can not insert pictures.

I only modify 'dataset.py' to make it load 'PLY' files directly. Besides that, everything is in default configuration. I also put the output file 'camel.ply' and 'cow.ply' as input to double-upsample and the result is right, so it may not be this reason.

`class PUNET_Dataset_Whole(torch_data.Dataset):
def init(self, data_dir='./datas/test_data/our_collected_data/test_ply'):
super().init()
file_list = os.listdir(data_dir)
self.names = [x.split('.')[0] for x in file_list]
self.sample_path = [os.path.join(data_dir, x) for x in file_list]

def __len__(self):
    return len(self.names)

def read_ply(self,filename):
	plydata = PlyData.read(filename)
	pc = plydata['vertex'].data
	pc_array = np.array([[x, y, z] for x,y,z in pc])
	return pc_array

def __getitem__(self, index):
    #points = np.loadtxt(self.sample_path[index])
    #print(self.sample_path[index])
    points = self.read_ply(self.sample_path[index])
    return points`

At first I think the reason is the data is too complicated so I sample 5k points in a small area and the result is still wrong. Right now I am really confused and your help is very needed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant