HorizonNav: Geometry-Guided Video Diffusion for Anticipatory 3D Robot Navigation
HorizonNav: Geometry-Guided Video Diffusion for Anticipatory 3D Robot Navigation
This paper presents HorizonNav, a video diffusion-based framework for anticipatory robot navigation that transforms a single RGB image and a text prompt into a geometrically aligned representation of plausible surrounding 3D space. The method integrates vision-language models, monocular ORB-SLAM3, and Video Depth Anything to improve geometric consistency in reconstructed environments, with experimental results showing increased planning feasibility and successful collision avoidance in simulated scenarios.
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
evaluable generated-camera trajectories
25
reconstructions from attempted cases
67 trials
View original evidence
A larger archive contains 58 usable reconstructions from 67 attempted cases
Open source S5reconstructions supporting a downstream plan
36
reconstruction error
≈0.3
reconstructions in corridor study
15
collision constraints violations
6
View original evidence
6 of 25 evaluable generated-camera trajectories violate the reconstructed collision constraints
Open source S5
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
th triangulated visual landmarks to constrain dense depth predictions across multiple views. This alignment improves the geometric consistency of the reconstructed 3D environment, addressing inconsistencies that temporal coherence of the generated video alone cannot resolve. The experimental evaluation of navigation in imagined 3D environment shows that across a 19-scene horizon study, planning-feasible scenes increa
Open source S4
se from 13/19 at P20 to 15/19 at P81, while successful RRT trials increase from 62/90 to 75/90. In a separate corridor study, 6 of 25 evaluable generated-camera trajectories violate the reconstructed collision constraints; RRT planner recovers feasible alternatives in 5 of those 6 cases. A larger archive contains 58 usable reconstructions from 67 attempted cases, of which 36 support a downstream plan and 22 do not, e
Open source S5
xposing reconstruction and free-space failure modes rather than hiding them without lidar. Across 15 corridor reconstructions, the mean reconstruction error is approximately 0.30 m at matched structural landmarks.
Open source S6
Source：arXiv Robotics — research abstracts · arxiv.org