4 lines
299 B
Markdown
4 lines
299 B
Markdown
# SpatialTrackerV2: Final version paper still polishing, ETA in one week.
|
|
|
|
## Overall
|
|
SpatialTrackerV2 proposes a end-to-end and differentiable pipeline to unify video depth, camera pose and 3D tracking. This unified pipeline enable large-scale joint training of both part in diverse types of data. |