ProCompNav: Proactive Instance Navigation with Comparative Judgment for Ambiguous User Queries
ProCompNav: Proactive Instance Navigation with Comparative Judgment for Ambiguous User Queries
This paper introduces ProCompNav, a framework for agentic robots to resolve ambiguous navigation requests through proactive clarification. The method employs a two-stage process involving candidate pool construction and Recursive Comparative Judgment (RCJ), which iteratively narrows options using binary questions based on attribute-value pairs. Experiments on interactive and non-interactive benchmarks show improved success rates and reduced response length compared to existing approaches, with human studies indicating user preference for the proposed interaction style.
The study presents ProCompNav, a two-stage framework for interactive instance navigation that uses Recursive Comparative Judgment to resolve ambiguous queries via binary questions, reporting improved success rates and reduced response length on benchmarks alongside user preference data from two human studies.
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.
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
t through Recursive Comparative Judgment (RCJ). RCJ iteratively narrows the pool by selecting an attribute-value pair that divides the candidates, asking the user a binary question, and removing inconsistent candidates, without requiring an attribute unique to the target. ProCompNav substantially improves Success Rate and reduces Response Length on an interactive navigation benchmark, and achieves the highest Success
Open source S4
Rate among evaluated methods on a non-interactive benchmark. In two human studies, participants preferred ProCompNav's binary questions to open-ended questions and its batched delivery to interleaved interaction. Our results highlight the need to jointly consider task performance and user-friendly interaction strategies when designing agentic robots.
Open source S5
at strongly matches the accumulated user-provided information. However, this process can require lengthy user responses and lead the agent to make premature decisions without distinguishing the target from distractors. To address these limitations, we propose Proactive Instance Navigation with Comparative Judgment (ProCompNav), a two-stage framework that first constructs a candidate pool and then identifies the targe
Open source S3
Source：arXiv Robotics — research abstracts · arxiv.org