SceneBind: Binding 'What' and 'Where' Across Vision, Audio, and Language

arXiv·medium signal

Existing omni-modal encoders capture instance-level semantics (what is present) but lack explicit 3D spatial structure (where it is). SceneBind represents each scene as a semantic-spatial entity, pairing a global semantic embedding with object-centric semantic-spatial slots across vision, audio, and language. Aimed at practitioners building embodied, robotics, or spatial-reasoning systems that need joint semantic and 3D grounding rather than flat multimodal similarity.