LiteNWM: Efficient Latent World Models for Onboard Visual Navigation in the Wild
LiteNWM: Efficient Latent World Models for Onboard Visual Navigation in the Wild
LiteNWM is a latent navigation world model that enables efficient visual navigation by sharing visual encoding across candidates and jointly predicting future representations. It achieves a 17.56% reduction in trajectory error and a 128.00-fold speedup on an RTX 5090 compared to NoMaD+NWM-XL.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
end-to-end speedup
128
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
predicts their action-conditioned future representations at multiple horizons, while a learned scorer uses these predictions to select trajectories. In offline evaluations on RECON, SCAND, and SACSoN, LiteNWM reduces macro-averaged trajectory error by 17.56% relative to NoMaD+NWM-XL and achieves a 128.00-fold end-to-end speedup on an RTX 5090. The same evaluator transfers from NoMaD to MBRA without proposer-specific
Open source S3
retraining, reducing MBRA's macro-averaged trajectory error by 16.2%. In real-robot experiments in unseen indoor and outdoor environments, LiteNWM improves navigation success from 43.3% to 83.3% relative to NoMaD. These results demonstrate that LiteNWM can be deployed for future-aware planning and closed-loop navigation on a physical robot.
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org