AI & ML
Racing the Clock: Teaching VLNs Budget-Conditioned Object Search
arXiv cs.RO0 views
arXiv:2611.17442v1 Announce Type: new Abstract: A rescue robot entering a building with thirty seconds on the clock should sprint toward the single most promising room; given ten minutes, it should instead explore a broader tour. Yet vision-language-navigation (VLN) models are typically evaluated under a fixed episode limit that the policy never observes, even though time-critical missions require behavior to adapt to the available budget. We study budget-conditioned object search: given a target category and an explicit action budget, an agent must find as many distinct target instances as possible before the budget expires. We investigate how to supervise VLNs for this task by comparing a privileged TSP expert, which plans near-optimal routes using the full map and true target locations, with a room-level Bellman expert that maximizes expected return under partial observability using high-level cross-scene semantic priors. Despite the TSP expert achieving substantially higher expert performance, its demonstrations are less effective for learning exploration: on held-out HM3D scenes, the Bellman-supervised VLN consistently outperforms the TSP-supervised VLN, with the advantage becoming more pronounced at longer budgets. At 2000 steps (500m), the Bellman-supervised VLN finds 71.4% of reachable target instances, compared with 65.5% for the TSP-supervised VLN and 44.8% for the pretrained VLN. These results highlight a distinction between optimal routing and optimal exploration: effective exploration supervision should optimize decisions under partial observability rather than imitate routes that are optimal only once the environment is fully known.
Source:https://arxiv.org/abs/2611.17442PreviousAgentic-Teacher: An Agentic Data Engine for Failure-Driven VLA Post-TrainingNextLearning Latent Goal Costs for Visual World-Model Planning with an Expanding Curriculum
Related posts
Denoising the US Census: Succinct Block Hierarchical Regression
0arXiv cs.CR
Adversarial Attacks on Downstream Weather Forecasting Models: Application to Tropical Cyclone Trajectory Prediction
0arXiv cs.CR
Cyber-Zero: Training Cybersecurity Agents without Runtime
0arXiv cs.CR
Coordinate Frame Recovery, Per Class NMS and Null Return Modelling for Sim2Real V2X LiDAR Detection
0arXiv cs.RO
Geometric Refinement and Distributional Alignment for Sim2Real Roadside LiDAR Detection
0arXiv cs.RO
D-VLA: A High-Concurrency Distributed Asynchronous Reinforcement Learning Framework for Vision-Language-Action Models
0arXiv cs.RO