HSfM jointly optimizes human meshes, dense scene pointmaps, and camera poses in a metric world frame, reducing world-frame human joint error from 3.5m to 1.0m on EgoHumans.
Key.net: Keypoint detection by hand- crafted and learned cnn filters
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Reconstructing People, Places, and Cameras
HSfM jointly optimizes human meshes, dense scene pointmaps, and camera poses in a metric world frame, reducing world-frame human joint error from 3.5m to 1.0m on EgoHumans.