output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre ...

Abstract: Expressing and controlling fine-grained spatial attributes of objects in large-scale models presents significant challenges, as these spatial attributes are often difficult to describe ...

Abstract: Endowing robots with the ability to understand natural language and execute grasping is a challenging task in a human-centric environment. Existing works on language-conditioned grasping ...

Generate Any Scene is a framework designed to systematically evaluate and improve text-to-vision models by generating a vast array of synthetic captions derived from dynamically constructed scene ...