DSCD-Nav: Dual-Stance Cooperative Debate for Object Navigation

TL;DR

This paper introduces DSCD-Nav, a novel decision mechanism for indoor object navigation that uses dual-stances and cooperative debate to improve reliability and efficiency in unfamiliar environments.

Contribution

It proposes a stance-based cooperative debate framework with evidence-aware arbitration to enhance decision-making in robot navigation under partial observability.

Findings

Improves success rates and path efficiency in indoor navigation tasks.

Reduces redundant exploration compared to existing methods.

Demonstrates effectiveness across multiple benchmark datasets.

Abstract

Adaptive navigation in unfamiliar indoor environments is crucial for household service robots. Despite advances in zero-shot perception and reasoning from vision-language models, existing navigation systems still rely on single-pass scoring at the decision layer, leading to overconfident long-horizon errors and redundant exploration. To tackle these problems, we propose Dual-Stance Cooperative Debate Navigation (DSCD-Nav), a decision mechanism that replaces one-shot scoring with stance-based cross-checking and evidence-aware arbitration to improve action reliability under partial observability. Specifically, given the same observation and candidate action set, we explicitly construct two stances by conditioning the evaluation on diverse and complementary objectives: a Task-Scene Understanding (TSU) stance that prioritizes goal progress from scene-layout cues, and a Safety-Information Balancing (SIB) stance that emphasizes risk and information value. The stances conduct a cooperative debate and make policy by cross-checking their top candidates with cue-grounded arguments. Then, a Navigation Consensus Arbitration (NCA) agent is employed to consolidate both sides' reasons and evidence, optionally triggering lightweight micro-probing to verify uncertain choices, preserving NCA's primary intent while disambiguating. Experiments on HM3Dv1, HM3Dv2, and MP3D demonstrate consistent improvements in success and path efficiency while reducing exploration redundancy.