ULN: Towards Underspecified Vision-and-Language Navigation
This addresses the issue of brittle VLN models in real-world scenarios with varied linguistic inputs, though it is incremental as it builds on existing VLN tasks.
The paper tackles the problem of Vision-and-Language Navigation (VLN) by introducing a new setting called Underspecified VLN (ULN) to handle multi-level underspecified instructions, which is more realistic than fine-grained ones, and proposes a framework with a classification module, navigation agent, and Exploitation-to-Exploration module that outperforms baselines by ~10% relative success rate.
Vision-and-Language Navigation (VLN) is a task to guide an embodied agent moving to a target position using language instructions. Despite the significant performance improvement, the wide use of fine-grained instructions fails to characterize more practical linguistic variations in reality. To fill in this gap, we introduce a new setting, namely Underspecified vision-and-Language Navigation (ULN), and associated evaluation datasets. ULN evaluates agents using multi-level underspecified instructions instead of purely fine-grained or coarse-grained, which is a more realistic and general setting. As a primary step toward ULN, we propose a VLN framework that consists of a classification module, a navigation agent, and an Exploitation-to-Exploration (E2E) module. Specifically, we propose to learn Granularity Specific Sub-networks (GSS) for the agent to ground multi-level instructions with minimal additional parameters. Then, our E2E module estimates grounding uncertainty and conducts multi-step lookahead exploration to improve the success rate further. Experimental results show that existing VLN models are still brittle to multi-level language underspecification. Our framework is more robust and outperforms the baselines on ULN by ~10% relative success rate across all levels.