Using translational optical flow and an uncertainty mask from forward-backward inconsistencies, a neural policy trained in differentiable simulation enables monocular FPV quadrotor flight at up to 11.79 m/s real-world speed with 93.3% success.
Flying in clutter on monocular rgb by learning in 3d radiance fields with domain adaptation
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.RO 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
GaussFly decouples representation learning from policy optimization via 3D Gaussian Splatting reconstruction and contrastive features to achieve superior sample efficiency and zero-shot sim-to-real transfer for AAV visuomotor policies.
citing papers explorer
-
Autonomous FPV Flight with Translational Optical Flow and Uncertainty Mask
Using translational optical flow and an uncertainty mask from forward-backward inconsistencies, a neural policy trained in differentiable simulation enables monocular FPV quadrotor flight at up to 11.79 m/s real-world speed with 93.3% success.
-
GaussFly: Contrastive Reinforcement Learning for Visuomotor Policies in 3D Gaussian Fields
GaussFly decouples representation learning from policy optimization via 3D Gaussian Splatting reconstruction and contrastive features to achieve superior sample efficiency and zero-shot sim-to-real transfer for AAV visuomotor policies.