Closed YhQIAO closed 11 months ago
gitouni
commented
12 months ago You don't need to worry about this. It's just a moderate warning from pykitti module.
The ground-truth poses of sequence 11-20 are not provided by KITTI because these sequences are used for fair testing, and that is why the pykitti.odometry() threw that warning. However, it does not affect the calibration process, so do not worry about that.
zzq123zz
commented
11 months ago how can i get the poses in the dataset?
gitouni
commented
11 months ago how can i get the poses in the dataset?
We all have no access to the poses of sequences for testing. They're used for online evaluation of odometry performance.
zzq123zz
commented
11 months ago How the poses is obtained in the figure below？
At 2023-09-13 17:46:03, "gitouni" @.***> wrote:
how can i get the poses in the dataset?
We all have no access to the poses of sequences for testing. They're used for online evaluation of odometry performance.
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>
gitouni
commented
11 months ago How the poses is obtained in the figure below？ At 2023-09-13 17:46:03, "gitouni" @.> wrote: how can i get the poses in the dataset? We all have no access to the poses of sequences for testing. They're used for online evaluation of odometry performance. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.>
Sorry, I cannot see that figure that you want to display. However, the poses are not necessary for this calibration, as CalibNet works on a single-frame mode.
zzq123zz
commented
11 months ago I have resulted! Thank you!
| @.*** | |
|---|---|
@.*** |
---- Replied Message ---- | From | @.> | | Date | 09/14/2023 09:34 | | To | @.> | | Cc | @.>@.> | | Subject | Re: [gitouni/CalibNet_pytorch] dataset question (Issue #7) |
How the poses is obtained in the figure below？ At 2023-09-13 17:46:03, "gitouni" @.> wrote: how can i get the poses in the dataset? We all have no access to the poses of sequences for testing. They're used for online evaluation of odometry performance. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.>
Sorry, I cannot see that figure that you want to display. However, the poses are not necessary for this calibration, as CalibNet works on a single-frame mode.
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>
why i got these messages: Ground truth poses are not avaialble for sequence 11. Ground truth poses are not avaialble for sequence 12. Ground truth poses are not avaialble for sequence 13. when i run test.py? it seems that kitti dataset dosen't have ground truth poses of sequence 11-22?