ROMA: AI model System for Real-World Object-Centric Multi-Sensory Active Perception

What happened

arXiv:2610.06955v1 Announce Type: new Abstract: Humans inherently understand the physical world through an active process. In stark contrast, existing multi-sensory robot systems mainly integrate sensory inputs rather than actively acquiring missing evidence through interactions.

ROMA integrates vision, audio, tactile, and force sensing into a reasoning-interaction-feedback loop. The model identifies missing evidence and determines the target objects, interactions, and modalities, while a physical interface executes the selected interactions and collects the multi-sensory feedback. Experiments show that ROMA can actively acquire (one company buying another) missing evidence and solve complex, long-chain multi-sensory perception tasks that existing methods struggle to handle, laying a strong perceptual foundation for active multi-sensory embodied agents.