1 paper
Hilton Raj, Vishnuram AV
In 3D environments, Embodied Agents answer spatially relevant questions through reasoning from a mixture of modalities including natural language, RGB images, point clouds, depth m…