DanielRoeder1

Rotated SLAM pose

1

In the readme it is outlined that the rotated SLAM pose is created to account for the slightly downward-facing camera angle (i.e 12 degrees). Looking at the code in the...

I am currently trying the network trained on the NYU dataset on a separate dataset (TUM RGBD desk see: https://vision.in.tum.de/data/datasets/rgbd-dataset/download#freiburg2_desk) I would have expected to get a similar performance but...

Custom learning rate scheduler affects TPU performance

2

## ❓ Questions and Help I have trained my transformer model once on a single GPU and once using a multi-core TPU. In both cases a batchsize of 256 is...

Camera Pose in world reference

1

I am trying to perform 3D reconstruction using the depth maps of the TUM datasets in conjunction with the poses of CCMSLAM. I know that the camera poses of CCMSLAM...

Understanding processing of Mind2Web dataset for Lumos grounding

2

Hello, I am trying to map the Lumos WebAgent grounding dataset onto the original Mind2Web dataset. Unfortunetly the ids (annotation_id, action_uid) were removed in the Lumos version but via query...