You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I tested the generate_test_tuples.py for SceneNet dataset (i converted it to the scannet dataset format so that i can use the scannet dataset class).
you suggest that frame-******.pose.txt must follow this form:
-0.384739 0.271466 -0.882203 4.98152
0.921157 0.0521417 -0.385682 1.46821
-0.0587002 -0.961035 -0.270124 1.51837
a (3,4) matrix.
But I am getting the following error:
scannet_dataset.py", line 579, in load_pose cam_T_world = np.linalg.inv(world_T_cam) numpy.linalg.LinAlgError: Last 2 dimensions of the array must be square
which to me indicates that frame-*****.pose.txt must be a square matrix (4,4).
Let me know if it makes sense.
Furthermore, do you think testing simplerecon on SceneNet dataset (input size: (320,240)) following Scannet dataset format is a reasonable approach?
Best,
Yohan
The text was updated successfully, but these errors were encountered:
I tested the
generate_test_tuples.py
for SceneNet dataset (i converted it to the scannet dataset format so that i can use the scannet dataset class).you suggest that frame-******.pose.txt must follow this form:
-0.384739 0.271466 -0.882203 4.98152
0.921157 0.0521417 -0.385682 1.46821
-0.0587002 -0.961035 -0.270124 1.51837
a (3,4) matrix.
But I am getting the following error:
scannet_dataset.py", line 579, in load_pose
cam_T_world = np.linalg.inv(world_T_cam)
numpy.linalg.LinAlgError: Last 2 dimensions of the array must be square
which to me indicates that frame-*****.pose.txt must be a square matrix (4,4).
Let me know if it makes sense.
Furthermore, do you think testing simplerecon on SceneNet dataset (input size: (320,240)) following Scannet dataset format is a reasonable approach?
Best,
Yohan
The text was updated successfully, but these errors were encountered: