LEGO learns a geometry-preserving latent space of humanoid designs from existing examples and performs gradient-free optimization in that space with task losses taken from human motion data.
A Geometric Perspective on Autoencoders
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LEGO: Latent-space Exploration for Geometry-aware Optimization of Humanoid Kinematic Design
LEGO learns a geometry-preserving latent space of humanoid designs from existing examples and performs gradient-free optimization in that space with task losses taken from human motion data.