Robotics Lab Home / Research Landscape / Spatial Intelligence

Robotics Research Landscape

Spatial Intelligence

Understanding and remembering 3D (and 4D) space: geometry, objects, scenes and language.

Spatial intelligence is the ability to represent, reason about and act in 3D space: where things are, how they are arranged and how they move.

Robots need scene representations that are metric, semantic, queryable and updatable, not just a sequence of images.

Foundation models bring semantic understanding, while neural and Gaussian representations bring dense geometry. Combining them is a core step towards embodied AI.

Loading Spatial Intelligence…