FAST-EQA: Efficient Embodied Question Answering with Global and Local Region Relevancy
2026-02-17 • Robotics
Robotics
AI summaryⓘ
The authors present FAST-EQA, a system designed to help a virtual agent answer questions about a scene by efficiently exploring only the relevant parts. It keeps a limited memory of possible targets in the environment and updates these guesses as it moves, which helps it handle questions about one or more objects without slowing down. The system also uses reasoning steps to make better decisions about where to look and what the answers might be. Compared to earlier methods, FAST-EQA runs faster while covering more of the scene and answering more accurately on several test benchmarks.
Embodied Question Answeringvisual scene understandingspatial reasoningChain-of-Thought reasoningnavigation policyexploration strategypartial observabilitymemory managementmulti-target questionsbenchmark evaluation
Authors
Haochen Zhang, Nirav Savaliya, Faizan Siddiqui, Enna Sachdeva
Abstract
Embodied Question Answering (EQA) combines visual scene understanding, goal-directed exploration, spatial and temporal reasoning under partial observability. A central challenge is to confine physical search to question-relevant subspaces while maintaining a compact, actionable memory of observations. Furthermore, for real-world deployment, fast inference time during exploration is crucial. We introduce FAST-EQA, a question-conditioned framework that (i) identifies likely visual targets, (ii) scores global regions of interest to guide navigation, and (iii) employs Chain-of-Thought (CoT) reasoning over visual memory to answer confidently. FAST-EQA maintains a bounded scene memory that stores a fixed-capacity set of region-target hypotheses and updates them online, enabling robust handling of both single and multi-target questions without unbounded growth. To expand coverage efficiently, a global exploration policy treats narrow openings and doors as high-value frontiers, complementing local target seeking with minimal computation. Together, these components focus the agent's attention, improve scene coverage, and improve answer reliability while running substantially faster than prior approaches. On HMEQA and EXPRESS-Bench, FAST-EQA achieves state-of-the-art performance, while performing competitively on OpenEQA and MT-HM3D.