Closed luoshuiyue closed 4 weeks ago
Hi,
This does not show where the error is. Can you type exprt HYDRA_FULL_ERROR=1
in you terminal and rerun the demo.py again?
I set HYDRA_FULL_ERROR=1,then run the command again:
Hi this seems your downloaded file is not complete.
Check this post for more details: https://stackoverflow.com/questions/71617570/pytorchstreamreader-failed-reading-zip-archive-failed-finding-central-directory
Hi @jytime thanks for the amazing work! I was getting a similar error. This is my o/p after setting export HYDRA_FULL_ERROR=1
. Also, seems like there was no convergence. I also get the following warning before the BA routine. Not sure if it's relevant but posting it as well. Thanks!
oh this seems all the frames are filtered out due to
Can you try to set filter_invalid_frame=False and see the results?
Thanks for your prompt reply! I don't get errors anymore but seems like there is still no convergence? For reference, I am feeding it 10 sequential images with a linearly moving camera with enough overlap between images i.e. the translation between two camera positions would be about 10 cm.
Hey @surabhi96
No convergence is okay here because it shows the optimisation result for last BA. A final cost of 0.18 seems not bad here. You can visualize the results by gr_visualize=True.
If it still does not work, feel free to send the images to me.
why I get this error when run demo.py:![1721632014677](https://github.com/user-attachments/assets/ebb09485-4385-4f68-ac54-8b98055bad83)