-
Notifications
You must be signed in to change notification settings - Fork 7
/
dataset.py
62 lines (52 loc) · 2.37 KB
/
dataset.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
# Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
#
# NVIDIA CORPORATION & AFFILIATES and its licensors retain all intellectual property
# and proprietary rights in and to this software, related documentation
# and any modifications thereto. Any use, reproduction, disclosure or
# distribution of this software and related documentation without an express
# license agreement from NVIDIA CORPORATION & AFFILIATES is strictly prohibited.
import os
import numpy as np
import h5py
import torch
class point_color_voxel_dataset(torch.utils.data.Dataset):
'''
This dataset returns:
1. Sampled points on the surfaces of 3D shapes;
2. The normals of the sampled points;
3. The RGB or RGBA colors of the sampled points;
4. Colored voxels of 3D shapes.
'''
def __init__(self, data_dir, point_batch_size, train):
self.data_dir = data_dir
self.point_batch_size = point_batch_size
self.train = train
obj_names = os.listdir(self.data_dir)
obj_names = sorted(obj_names)
if self.train is None:
self.start_idx = 0
self.obj_names = obj_names
print("Total#", "all", len(self.obj_names))
elif self.train:
self.start_idx = 0
self.obj_names = obj_names[:int(len(obj_names)*0.8)]
print("Total#", "train", len(self.obj_names))
else:
self.start_idx = int(len(obj_names)*0.8)
self.obj_names = obj_names[int(len(obj_names)*0.8):]
print("Total#", "test", len(self.obj_names))
def __len__(self):
return len(self.obj_names)
def __getitem__(self, index):
hdf5_dir = self.data_dir+"/"+self.obj_names[index]+"/vertices_normals_colors_voxels.hdf5"
grid_size = 64
hdf5_file = h5py.File(hdf5_dir, 'r')
rand_idcs = np.random.randint(len(hdf5_file["vertices"])-self.point_batch_size+1)
vertices = hdf5_file["vertices"][rand_idcs:rand_idcs+self.point_batch_size]
normals = hdf5_file["normals"][rand_idcs:rand_idcs+self.point_batch_size]
colors = hdf5_file["colors"][rand_idcs:rand_idcs+self.point_batch_size]
voxels = hdf5_file["voxel_color"][:]
hdf5_file.close()
colors = colors[:,:3] #RGB only, remove alpha
voxels = np.transpose(voxels, (3,0,1,2)).astype(np.float32)
return vertices, normals, colors, voxels