Grounding 3D Object Affordance with Language Instructions, Visual Observations and Interactions | Read Paper on Bytez