Active Test-time Vision-Language Navigation

TL;DR

This paper introduces ATENA, a test-time active learning framework for vision-language navigation that incorporates human feedback and mixture entropy optimization to improve navigation accuracy in unfamiliar environments.

Contribution

ATENA is the first to combine active learning, mixture entropy optimization, and human-in-the-loop feedback for test-time vision-language navigation.

Findings

ATENA outperforms baseline methods on REVERIE, R2R, and R2R-CE benchmarks.

It effectively handles distributional shifts during test time.

The framework improves uncertainty calibration and decision-making in navigation tasks.

Abstract

Vision-Language Navigation (VLN) policies trained on offline datasets often exhibit degraded task performance when deployed in unfamiliar navigation environments at test time, where agents are typically evaluated without access to external interaction or feedback. Entropy minimization has emerged as a practical solution for reducing prediction uncertainty at test time; however, it can suffer from accumulated errors, as agents may become overconfident in incorrect actions without sufficient contextual grounding. To tackle these challenges, we introduce ATENA (Active TEst-time Navigation Agent), a test-time active learning framework that enables a practical human-robot interaction via episodic feedback on uncertain navigation outcomes. In particular, ATENA learns to increase certainty in successful episodes and decrease it in failed ones, improving uncertainty calibration. Here, we propose mixture entropy optimization, where entropy is obtained from a combination of the action and pseudo-expert distributions-a hypothetical action distribution assuming the agent's selected action to be optimal-controlling both prediction confidence and action preference. In addition, we propose a self-active learning strategy that enables an agent to evaluate its navigation outcomes based on confident predictions. As a result, the agent stays actively engaged throughout all iterations, leading to well-grounded and adaptive decision-making. Extensive evaluations on challenging VLN benchmarks-REVERIE, R2R, and R2R-CE-demonstrate that ATENA successfully overcomes distributional shifts at test time, outperforming the compared baseline methods across various settings.