Skip to content

GitLab

Unverified Commit 4eed122d authored by Yezhen Cong's avatar Yezhen Cong Committed by GitHub
Browse files

[Feature] Support ImVoteNet complete model (#352) 



* Added image loading in SUNRGB-D dataset (#195)

* image loading

* format and docstring fix

* removed irrelevant files

* removed irrelevant files

* load image only if modality is pc+img

* added modality like nuscenes

* Added imvotenet image branch pretrain (#217)

* image loading

* naive commit

* format and docstring fix

* removed irrelevant files

* removed irrelevant files

* load image only if modality is pc+img

* added modality like nuscenes

* pretrain_2d_model

* finetune sunrgbd

* finetune sunrgbd

* deleted unused code

* fixed a bug

* resolve conflict

* update config file

* fix docstring and configs

* integrated vote fusion

* coords transform and unit test

* Update docstring

* refactor and add unit test

* fix bug caused by mmcv upgrade; delete pdb breakpoint

* add point fusion unittest

* remove unused file

* fix typos

* updates

* add assertion info

* update

* add unittest

* add vote fusion unittest

* add vote fusion unittest

* [Refactor] VoteNet refactor (#322)

* votenet refactor

* remove file

* minor update

* docstring

* initial update of imvotenet

* [Feature] Support vote fusion (#297)

* integrated vote fusion

* coords transform and unit test

* Update docstring

* refactor and add unit test

* add point fusion unittest

* remove unused file

* updates

* add assertion info

* update

* add unittest

* add vote fusion unittest

* add vote fusion unittest

* minor update

* docstring

* change np ops to torch

* refactor test

* update

* refactor of image mlp and np random ops to torch

* add docstring

* add config and mod dataset

* fix bugs

* add_comments

* fix bugs

* fix_bug

* fix bug

* fix bug

* fix bug

* fix bug

* final fix

* fix bug

* ?

* add docstring

* move train/test cfg

* change img mlp default param

* rename config

* minor mod

* change config name

* move train/test cfg

* some fixes and 2d utils

* fix config name

* fix config override issue

* config simplify & reformat

* explicitly set eval mode->override train()

* add fix_img_branch to config

* remove set_img_branch_eval_mode

* temporal fix, change calibs to calib

* more docstring and view/reshape, expand/repeat change

* complete imvotenet docstring

* fix docstring

* add config and some minor fix

* rename config
Co-authored-by: default avatarZwwWayne <wayne.zw@outlook.com>
parent 097b66ee
Show whitespace changes
Inline Side-by-side
Showing with 594 additions and 12 deletions
mmdet3d/models/utils/__init__.py
View file @ 4eed122d
from .clip_sigmoid import clip_sigmoid
from .mlp import MLP
__all__ = ['clip_sigmoid']
__all__ = ['clip_sigmoid', 'MLP']
mmdet3d/models/utils/mlp.py 0 → 100644
View file @ 4eed122d
from mmcv.cnn import ConvModule
from torch import nn as nn
class MLP(nn.Module):
"""A simple MLP module.
Pass features (B, C, N) through an MLP.
Args:
in_channels (int): Number of channels of input features.
Default: 18.
conv_channels (tuple[int]): Out channels of the convolution.
Default: (256, 256).
conv_cfg (dict): Config of convolution.
Default: dict(type='Conv1d').
norm_cfg (dict): Config of normalization.
Default: dict(type='BN1d').
act_cfg (dict): Config of activation.
Default: dict(type='ReLU').
"""
def __init__(self,
in_channel=18,
conv_channels=(256, 256),
conv_cfg=dict(type='Conv1d'),
norm_cfg=dict(type='BN1d'),
act_cfg=dict(type='ReLU')):
super().__init__()
self.mlp = nn.Sequential()
prev_channels = in_channel
for i, conv_channel in enumerate(conv_channels):
self.mlp.add_module(
f'layer{i}',
ConvModule(
prev_channels,
conv_channels[i],
1,
padding=0,
conv_cfg=conv_cfg,
norm_cfg=norm_cfg,
act_cfg=act_cfg,
bias=True,
inplace=True))
prev_channels = conv_channels[i]
def forward(self, img_features):
return self.mlp(img_features)
tests/test_models/test_fusion/test_fusion_coord_trans.py 0 → 100644
View file @ 4eed122d
"""Tests coords transformation in fusion modules.
CommandLine:
pytest tests/test_models/test_fusion/test_fusion_coord_trans.py
"""
import torch
from mmdet3d.models.fusion_layers import apply_3d_transformation
def test_coords_transformation():
"""Test the transformation of 3d coords."""
# H+R+S+T, not reverse, depth
img_meta = {
'pcd_scale_factor':
1.2311e+00,
'pcd_rotation': [[8.660254e-01, 0.5, 0], [-0.5, 8.660254e-01, 0],
[0, 0, 1.0e+00]],
'pcd_trans': [1.111e-02, -8.88e-03, 0.0],
'pcd_horizontal_flip':
True,
'transformation_3d_flow': ['HF', 'R', 'S', 'T']
}
pcd = torch.tensor([[-5.2422e+00, -2.9757e-01, 4.0021e+01],
[-9.1435e-01, 2.6675e+01, -5.5950e+00],
[2.0089e-01, 5.8098e+00, -3.5409e+01],
[-1.9461e-01, 3.1309e+01, -1.0901e+00]])
pcd_transformed = apply_3d_transformation(
pcd, 'DEPTH', img_meta, reverse=False)
expected_tensor = torch.tensor(
[[5.78332345e+00, 2.900697e+00, 4.92698531e+01],
[-1.5433839e+01, 2.8993850e+01, -6.8880045e+00],
[-3.77929405e+00, 6.061661e+00, -4.35920199e+01],
[-1.9053658e+01, 3.3491436e+01, -1.34202211e+00]])
assert torch.allclose(expected_tensor, pcd_transformed, 1e-4)
# H+R+S+T, reverse, depth
img_meta = {
'pcd_scale_factor':
7.07106781e-01,
'pcd_rotation': [[7.07106781e-01, 7.07106781e-01, 0.0],
[-7.07106781e-01, 7.07106781e-01, 0.0],
[0.0, 0.0, 1.0e+00]],
'pcd_trans': [0.0, 0.0, 0.0],
'pcd_horizontal_flip':
False,
'transformation_3d_flow': ['HF', 'R', 'S', 'T']
}
pcd = torch.tensor([[-5.2422e+00, -2.9757e-01, 4.0021e+01],
[-9.1435e+01, 2.6675e+01, -5.5950e+00],
[6.061661e+00, -0.0, -1.0e+02]])
pcd_transformed = apply_3d_transformation(
pcd, 'DEPTH', img_meta, reverse=True)
expected_tensor = torch.tensor(
[[-5.53977e+00, 4.94463e+00, 5.65982409e+01],
[-6.476e+01, 1.1811e+02, -7.91252488e+00],
[6.061661e+00, -6.061661e+00, -1.41421356e+02]])
assert torch.allclose(expected_tensor, pcd_transformed, 1e-4)
# H+R+S+T, not reverse, camera
img_meta = {
'pcd_scale_factor':
1.0 / 7.07106781e-01,
'pcd_rotation': [[7.07106781e-01, 0.0, 7.07106781e-01],
[0.0, 1.0e+00, 0.0],
[-7.07106781e-01, 0.0, 7.07106781e-01]],
'pcd_trans': [1.0e+00, -1.0e+00, 0.0],
'pcd_horizontal_flip':
True,
'transformation_3d_flow': ['HF', 'S', 'R', 'T']
}
pcd = torch.tensor([[-5.2422e+00, 4.0021e+01, -2.9757e-01],
[-9.1435e+01, -5.5950e+00, 2.6675e+01],
[6.061661e+00, -1.0e+02, -0.0]])
pcd_transformed = apply_3d_transformation(
pcd, 'CAMERA', img_meta, reverse=False)
expected_tensor = torch.tensor(
[[6.53977e+00, 5.55982409e+01, 4.94463e+00],
[6.576e+01, -8.91252488e+00, 1.1811e+02],
[-5.061661e+00, -1.42421356e+02, -6.061661e+00]])
assert torch.allclose(expected_tensor, pcd_transformed, 1e-4)
# V, reverse, camera
img_meta = {'pcd_vertical_flip': True, 'transformation_3d_flow': ['VF']}
pcd_transformed = apply_3d_transformation(
pcd, 'CAMERA', img_meta, reverse=True)
expected_tensor = torch.tensor([[-5.2422e+00, 4.0021e+01, 2.9757e-01],
[-9.1435e+01, -5.5950e+00, -2.6675e+01],
[6.061661e+00, -1.0e+02, 0.0]])
assert torch.allclose(expected_tensor, pcd_transformed, 1e-4)
# V+H, not reverse, depth
img_meta = {
'pcd_vertical_flip': True,
'pcd_horizontal_flip': True,
'transformation_3d_flow': ['VF', 'HF']
}
pcd_transformed = apply_3d_transformation(
pcd, 'DEPTH', img_meta, reverse=False)
expected_tensor = torch.tensor([[5.2422e+00, -4.0021e+01, -2.9757e-01],
[9.1435e+01, 5.5950e+00, 2.6675e+01],
[-6.061661e+00, 1.0e+02, 0.0]])
assert torch.allclose(expected_tensor, pcd_transformed, 1e-4)
# V+H, reverse, lidar
img_meta = {
'pcd_vertical_flip': True,
'pcd_horizontal_flip': True,
'transformation_3d_flow': ['VF', 'HF']
}
pcd_transformed = apply_3d_transformation(
pcd, 'LIDAR', img_meta, reverse=True)
expected_tensor = torch.tensor([[5.2422e+00, -4.0021e+01, -2.9757e-01],
[9.1435e+01, 5.5950e+00, 2.6675e+01],
[-6.061661e+00, 1.0e+02, 0.0]])
assert torch.allclose(expected_tensor, pcd_transformed, 1e-4)
tests/test_models/test_fusion/test_point_fusion.py 0 → 100644
View file @ 4eed122d
"""Tests the core function of point fusion.
CommandLine:
pytest tests/test_models/test_fusion/test_point_fusion.py
"""
import torch
from mmdet3d.models.fusion_layers import PointFusion
def test_sample_single():
# this function makes sure the rewriting of 3d coords transformation
# in point fusion does not change the original behaviour
lidar2img = torch.tensor(
[[6.0294e+02, -7.0791e+02, -1.2275e+01, -1.7094e+02],
[1.7678e+02, 8.8088e+00, -7.0794e+02, -1.0257e+02],
[9.9998e-01, -1.5283e-03, -5.2907e-03, -3.2757e-01],
[0.0000e+00, 0.0000e+00, 0.0000e+00, 1.0000e+00]])
# all use default
img_meta = {
'transformation_3d_flow': ['R', 'S', 'T', 'HF'],
'input_shape': [370, 1224],
'img_shape': [370, 1224],
'lidar2img': lidar2img,
}
# dummy parameters
fuse = PointFusion(1, 1, 1, 1)
img_feat = torch.arange(370 * 1224)[None, ...].view(
370, 1224)[None, None, ...].float() / (370 * 1224)
pts = torch.tensor([[8.356, -4.312, -0.445], [11.777, -6.724, -0.564],
[6.453, 2.53, -1.612], [6.227, -3.839, -0.563]])
out = fuse.sample_single(img_feat, pts, img_meta)
expected_tensor = torch.tensor(
[0.5560822, 0.5476625, 0.9687978, 0.6241757])
assert torch.allclose(expected_tensor, out, 1e-4)
pcd_rotation = torch.tensor([[8.660254e-01, 0.5, 0],
[-0.5, 8.660254e-01, 0], [0, 0, 1.0e+00]])
pcd_scale_factor = 1.111
pcd_trans = torch.tensor([1.0, -1.0, 0.5])
pts = pts @ pcd_rotation
pts *= pcd_scale_factor
pts += pcd_trans
pts[:, 1] = -pts[:, 1]
# not use default
img_meta.update({
'pcd_scale_factor': pcd_scale_factor,
'pcd_rotation': pcd_rotation,
'pcd_trans': pcd_trans,
'pcd_horizontal_flip': True
})
out = fuse.sample_single(img_feat, pts, img_meta)
expected_tensor = torch.tensor(
[0.5560822, 0.5476625, 0.9687978, 0.6241757])
assert torch.allclose(expected_tensor, out, 1e-4)
tests/test_models/test_fusion/test_vote_fusion.py 0 → 100644
View file @ 4eed122d
"""Tests the core function of vote fusion.
CommandLine:
pytest tests/test_models/test_fusion/test_vote_fusion.py
"""
import torch
from mmdet3d.models.fusion_layers import VoteFusion
def test_vote_fusion():
img_meta = {
'ori_shape': (530, 730, 3),
'img_shape': (600, 826, 3),
'pad_shape': (608, 832, 3),
'scale_factor':
torch.tensor([1.1315, 1.1321, 1.1315, 1.1321]),
'flip':
False,
'pcd_horizontal_flip':
False,
'pcd_vertical_flip':
False,
'pcd_trans':
torch.tensor([0., 0., 0.]),
'pcd_scale_factor':
1.0308290128214932,
'pcd_rotation':
torch.tensor([[0.9747, 0.2234, 0.0000], [-0.2234, 0.9747, 0.0000],
[0.0000, 0.0000, 1.0000]]),
'transformation_3d_flow': ['HF', 'R', 'S', 'T']
}
calibs = {
'Rt':
torch.tensor([[[0.979570, 0.047954, -0.195330],
[0.047954, 0.887470, 0.458370],
[0.195330, -0.458370, 0.867030]]]),
'K':
torch.tensor([[[529.5000, 0.0000, 365.0000],
[0.0000, 529.5000, 265.0000], [0.0000, 0.0000,
1.0000]]])
}
bboxes = torch.tensor([[[
5.4286e+02, 9.8283e+01, 6.1700e+02, 1.6742e+02, 9.7922e-01, 3.0000e+00
], [
4.2613e+02, 8.4646e+01, 4.9091e+02, 1.6237e+02, 9.7848e-01, 3.0000e+00
], [
2.5606e+02, 7.3244e+01, 3.7883e+02, 1.8471e+02, 9.7317e-01, 3.0000e+00
], [
6.0104e+02, 1.0648e+02, 6.6757e+02, 1.9216e+02, 8.4607e-01, 3.0000e+00
], [
2.2923e+02, 1.4984e+02, 7.0163e+02, 4.6537e+02, 3.5719e-01, 0.0000e+00
], [
2.5614e+02, 7.4965e+01, 3.3275e+02, 1.5908e+02, 2.8688e-01, 3.0000e+00
], [
9.8718e+00, 1.4142e+02, 2.0213e+02, 3.3878e+02, 1.0935e-01, 3.0000e+00
], [
6.1930e+02, 1.1768e+02, 6.8505e+02, 2.0318e+02, 1.0720e-01, 3.0000e+00
]]])
seeds_3d = torch.tensor([[[0.044544, 1.675476, -1.531831],
[2.500625, 7.238662, -0.737675],
[-0.600003, 4.827733, -0.084022],
[1.396212, 3.994484, -1.551180],
[-2.054746, 2.012759, -0.357472],
[-0.582477, 6.580470, -1.466052],
[1.313331, 5.722039, 0.123904],
[-1.107057, 3.450359, -1.043422],
[1.759746, 5.655951, -1.519564],
[-0.203003, 6.453243, 0.137703],
[-0.910429, 0.904407, -0.512307],
[0.434049, 3.032374, -0.763842],
[1.438146, 2.289263, -1.546332],
[0.575622, 5.041906, -0.891143],
[-1.675931, 1.417597, -1.588347]]])
imgs = torch.linspace(
-1, 1, steps=608 * 832).reshape(1, 608, 832).repeat(3, 1, 1)[None]
expected_tensor1 = torch.tensor(
[[[
0.000000e+00, -0.000000e+00, 0.000000e+00, -0.000000e+00,
0.000000e+00, 1.193706e-01, -0.000000e+00, -2.879214e-01,
-0.000000e+00, 0.000000e+00, 1.422463e-01, -6.474612e-01,
-0.000000e+00, 1.490057e-02, 0.000000e+00
],
[
0.000000e+00, -0.000000e+00, -0.000000e+00, 0.000000e+00,
0.000000e+00, -1.873745e+00, -0.000000e+00, 1.576240e-01,
0.000000e+00, -0.000000e+00, -3.646177e-02, -7.751858e-01,
0.000000e+00, 9.593642e-02, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, -6.263277e-02, 0.000000e+00, -3.646387e-01,
0.000000e+00, 0.000000e+00, -5.875812e-01, -6.263450e-02,
0.000000e+00, 1.149264e-01, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 8.899736e-01, 0.000000e+00, 9.019017e-01,
0.000000e+00, 0.000000e+00, 6.917775e-01, 8.899733e-01,
0.000000e+00, 9.812444e-01, 0.000000e+00
],
[
-0.000000e+00, -0.000000e+00, -0.000000e+00, -0.000000e+00,
-0.000000e+00, -4.516903e-01, -0.000000e+00, -2.315422e-01,
-0.000000e+00, -0.000000e+00, -4.197519e-01, -4.516906e-01,
-0.000000e+00, -1.547615e-01, -0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 3.571937e-01, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 3.571937e-01,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 9.731653e-01,
0.000000e+00, 0.000000e+00, 1.093455e-01, 0.000000e+00,
0.000000e+00, 8.460656e-01, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
2.316288e-03, -1.948284e-03, -3.694394e-03, 2.176163e-04,
-3.882605e-03, -1.901490e-03, -3.355042e-03, -1.774631e-03,
-6.981542e-04, -3.886823e-03, -1.302233e-03, -1.189933e-03,
2.540967e-03, -1.834944e-03, 1.032048e-03
],
[
2.316288e-03, -1.948284e-03, -3.694394e-03, 2.176163e-04,
-3.882605e-03, -1.901490e-03, -3.355042e-03, -1.774631e-03,
-6.981542e-04, -3.886823e-03, -1.302233e-03, -1.189933e-03,
2.540967e-03, -1.834944e-03, 1.032048e-03
],
[
2.316288e-03, -1.948284e-03, -3.694394e-03, 2.176163e-04,
-3.882605e-03, -1.901490e-03, -3.355042e-03, -1.774631e-03,
-6.981542e-04, -3.886823e-03, -1.302233e-03, -1.189933e-03,
2.540967e-03, -1.834944e-03, 1.032048e-03
]]])
expected_tensor2 = torch.tensor([[
False, False, False, False, False, True, False, True, False, False,
True, True, False, True, False, False, False, False, False, False,
False, False, True, False, False, False, False, False, True, False,
False, False, False, False, False, False, False, False, False, False,
False, False, False, True, False
]])
expected_tensor3 = torch.tensor(
[[[
-0.000000e+00, -0.000000e+00, -0.000000e+00, -0.000000e+00,
0.000000e+00, -0.000000e+00, -0.000000e+00, 0.000000e+00,
-0.000000e+00, -0.000000e+00, 0.000000e+00, -0.000000e+00,
-0.000000e+00, 1.720988e-01, 0.000000e+00
],
[
0.000000e+00, -0.000000e+00, -0.000000e+00, 0.000000e+00,
-0.000000e+00, 0.000000e+00, -0.000000e+00, 0.000000e+00,
0.000000e+00, -0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 4.824460e-02, 0.000000e+00
],
[
-0.000000e+00, -0.000000e+00, -0.000000e+00, -0.000000e+00,
-0.000000e+00, -0.000000e+00, -0.000000e+00, 0.000000e+00,
-0.000000e+00, -0.000000e+00, -0.000000e+00, -0.000000e+00,
-0.000000e+00, 1.447314e-01, -0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 9.759269e-01, 0.000000e+00
],
[
-0.000000e+00, -0.000000e+00, -0.000000e+00, -0.000000e+00,
-0.000000e+00, -0.000000e+00, -0.000000e+00, -0.000000e+00,
-0.000000e+00, -0.000000e+00, -0.000000e+00, -0.000000e+00,
-0.000000e+00, -1.631542e-01, -0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 1.072001e-01, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00, 0.000000e+00,
0.000000e+00, 0.000000e+00, 0.000000e+00
],
[
2.316288e-03, -1.948284e-03, -3.694394e-03, 2.176163e-04,
-3.882605e-03, -1.901490e-03, -3.355042e-03, -1.774631e-03,
-6.981542e-04, -3.886823e-03, -1.302233e-03, -1.189933e-03,
2.540967e-03, -1.834944e-03, 1.032048e-03
],
[
2.316288e-03, -1.948284e-03, -3.694394e-03, 2.176163e-04,
-3.882605e-03, -1.901490e-03, -3.355042e-03, -1.774631e-03,
-6.981542e-04, -3.886823e-03, -1.302233e-03, -1.189933e-03,
2.540967e-03, -1.834944e-03, 1.032048e-03
],
[
2.316288e-03, -1.948284e-03, -3.694394e-03, 2.176163e-04,
-3.882605e-03, -1.901490e-03, -3.355042e-03, -1.774631e-03,
-6.981542e-04, -3.886823e-03, -1.302233e-03, -1.189933e-03,
2.540967e-03, -1.834944e-03, 1.032048e-03
]]])
fusion = VoteFusion()
out1, out2 = fusion(imgs, bboxes, seeds_3d, [img_meta], calibs)
assert torch.allclose(expected_tensor1, out1[:, :, :15], 1e-3)
assert torch.allclose(expected_tensor2.float(), out2.float(), 1e-3)
assert torch.allclose(expected_tensor3, out1[:, :, 30:45], 1e-3)
out1, out2 = fusion(imgs, bboxes[:, :2], seeds_3d, [img_meta], calibs)
out1 = out1[:, :15, 30:45]
out2 = out2[:, 30:45].float()
assert torch.allclose(torch.zeros_like(out1), out1, 1e-3)
assert torch.allclose(torch.zeros_like(out2), out2, 1e-3)
tests/test_utils/test_coord_3d_mode.py
View file @ 4eed122d
......@@ -62,21 +62,21 @@ def test_points_conversion():
convert_depth_points = cam_points.convert_to(Coord3DMode.DEPTH)
expected_tensor = torch.tensor([[
-5.2422e+00, -2.9757e-01, 4.0021e+01, 6.6660e-01, 1.9560e-01,
-5.2422e+00, 2.9757e-01, -4.0021e+01, 6.6660e-01, 1.9560e-01,
4.9740e-01, 9.4090e-01
],
[
-2.6675e+01, 9.1435e-01, 5.5950e+00,
-2.6675e+01, -9.1435e-01, -5.5950e+00,
1.5020e-01, 3.7070e-01, 1.0860e-01,
6.2970e-01
],
[
-5.8098e+00, -2.0089e-01, 3.5409e+01,
-5.8098e+00, 2.0089e-01, -3.5409e+01,
6.5650e-01, 6.2480e-01, 6.9540e-01,
2.5380e-01
],
[
-3.1309e+01, 1.9461e-01, 1.0901e+00,
-3.1309e+01, -1.9461e-01, -1.0901e+00,
2.8030e-01, 2.5800e-02, 4.8960e-01,
3.2690e-01
]])
......@@ -157,21 +157,21 @@ def test_points_conversion():
convert_cam_points = depth_points.convert_to(Coord3DMode.CAM)
expected_tensor = torch.tensor([[
-5.2422e+00, 2.9757e-01, -4.0021e+01, 6.6660e-01, 1.9560e-01,
-5.2422e+00, -2.9757e-01, 4.0021e+01, 6.6660e-01, 1.9560e-01,
4.9740e-01, 9.4090e-01
],
[
-2.6675e+01, -9.1435e-01, -5.5950e+00,
-2.6675e+01, 9.1435e-01, 5.5950e+00,
1.5020e-01, 3.7070e-01, 1.0860e-01,
6.2970e-01
],
[
-5.8098e+00, 2.0089e-01, -3.5409e+01,
-5.8098e+00, -2.0089e-01, 3.5409e+01,
6.5650e-01, 6.2480e-01, 6.9540e-01,
2.5380e-01
],
[
-3.1309e+01, -1.9461e-01, -1.0901e+00,
-3.1309e+01, 1.9461e-01, 1.0901e+00,
2.8030e-01, 2.5800e-02, 4.8960e-01,
3.2690e-01
]])
......@@ -182,6 +182,22 @@ def test_points_conversion():
assert torch.allclose(expected_tensor, convert_cam_points.tensor, 1e-4)
assert torch.allclose(cam_point_tensor, convert_cam_points.tensor, 1e-4)
rt_mat_provided = torch.tensor([[0.99789, -0.012698, -0.063678],
[-0.012698, 0.92359, -0.38316],
[0.063678, 0.38316, 0.92148]])
depth_points_new = torch.cat([
depth_points.tensor[:, :3] @ rt_mat_provided.t(),
depth_points.tensor[:, 3:]
],
dim=1)
cam_point_tensor_new = Coord3DMode.convert_point(
depth_points_new,
Coord3DMode.DEPTH,
Coord3DMode.CAM,
rt_mat=rt_mat_provided)
assert torch.allclose(expected_tensor, cam_point_tensor_new, 1e-4)
convert_lidar_points = depth_points.convert_to(Coord3DMode.LIDAR)
expected_tensor = torch.tensor([[
4.0021e+01, 5.2422e+00, 2.9757e-01, 6.6660e-01, 1.9560e-01, 4.9740e-01,
......
tools/data_converter/sunrgbd_data_utils.py
View file @ 4eed122d
......@@ -111,8 +111,9 @@ class SUNRGBDData(object):
calib_filepath = osp.join(self.calib_dir, f'{idx:06d}.txt')
lines = [line.rstrip() for line in open(calib_filepath)]
Rt = np.array([float(x) for x in lines[0].split(' ')])
Rt = np.reshape(Rt, (3, 3), order='F')
Rt = np.reshape(Rt, (3, 3), order='F').astype(np.float32)
K = np.array([float(x) for x in lines[1].split(' ')])
K = np.reshape(Rt, (3, 3), order='F').astype(np.float32)
return K, Rt
def get_label_objects(self, idx):
......@@ -155,8 +156,7 @@ class SUNRGBDData(object):
osp.join(self.root_dir, 'points', f'{sample_idx:06d}.bin'))
info['pts_path'] = osp.join('points', f'{sample_idx:06d}.bin')
img_name = osp.join(self.image_dir, f'{sample_idx:06d}')
img_path = osp.join(self.image_dir, img_name)
img_path = osp.join('image', f'{sample_idx:06d}.jpg')
image_info = {
'image_idx': sample_idx,
'image_shape': self.get_image_shape(sample_idx),
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment