Visual navigation improves by predicting free space from past views
EchoNav: Evidential BEV Completion from Historical Observations for Visual Navigation
Robotics
Summary
Robots moving in cluttered spaces need to figure out what’s around them, including things they can’t see right now. The authors developed a method that uses recent camera images to make a map from above, showing which areas are open or blocked. Their system can guess both observed and hidden obstacles with confidence, all without needing extra sensors or pose information. Tests in simulations and real robots show it works much better than previous methods for reaching navigation goals.
What this means in practice
- •For robotics developers: Improve navigation in cluttered environments by generating confident local maps from short camera sequences without depth sensors or pose input.
- •For mobile robot operators: Use accurate occupancy predictions to increase goal-reaching success rates during indoor navigation in visually complex scenes.
Authors
Diwen Liu, Yiyang Shao, Borui Wang, Chi Zhang, Yunhui Liu
Abstract
Visual navigation in cluttered environments requires robots to recover previously observed structure and infer obstacles and free space that remain occluded. We present a framework that combines temporal aggregation and occupancy completion in an evidential bird's-eye-view (BEV) representation predicted from short RGB sequences. A decoder adapts frozen VGGT-Omega features and combines historical residual proposals using cellwise weights and a learned no-update option. The BEV predictor produces a local map in the latest ego frame without depth or camera-pose inputs or explicit geometric warping. The evidential model treats observed free space separately and predicts occupancy with associated confidence across the remaining supported region. We train the predictor on simulated RGB sequences paired with complete and visibility-masked occupancy annotations. Our experiments evaluate navigation in simulation and on a wheeled robot, with additional demonstrations on quadruped and humanoid robots. In simulated sequential two-goal navigation, our system achieves 89.6% success, outperforming the strongest evaluated baseline by 32.5 percentage points. Across 60 wheeled-robot trials in 20 scenes, it achieves 73.3% success, compared with 33.3% for the strongest evaluated baseline.