I want to do Lapida dataset with the cubemap slam code currently running on mp4. So when I convert the lafida dataset into an mp4 video and run it on that slam, the slam is executed, but it's not the point cloud creation part or tracking part at all. I've set all the resolution and fps of the video the same, what's the problem?
I thought that if it had the same mp4 format, the cubemap slam would be executed the same, so I used the cubemap slam by converting the individual images of the lafida dataset into mp4 with the same fps and resolution using the tool. In that case, the cubemapslam runs, but it is not creating a point cloud at all. So of course, tracking is not working at all. I think the existing mp4 video had a cubemap-type black border in the video, but the newly created mp4 does not have that border, so I think that's the problem.