HUMANISE: Language-conditioned Human Motion Generation in 3D Scenes
This work addresses the problem of limited and low-quality datasets for human-scene interaction in computer vision and robotics, enabling new applications in animation and simulation, though it is incremental by building on existing motion generation methods.
The authors tackled the challenge of generating diverse, scene-aware, and goal-oriented human motions in 3D scenes by creating a large-scale synthetic dataset called HUMANISE, which aligns human motions with 3D scenes and language descriptions, and they developed a model that produces semantically consistent motions, as demonstrated in experiments.
Learning to generate diverse scene-aware and goal-oriented human motions in 3D scenes remains challenging due to the mediocre characteristics of the existing datasets on Human-Scene Interaction (HSI); they only have limited scale/quality and lack semantics. To fill in the gap, we propose a large-scale and semantic-rich synthetic HSI dataset, denoted as HUMANISE, by aligning the captured human motion sequences with various 3D indoor scenes. We automatically annotate the aligned motions with language descriptions that depict the action and the unique interacting objects in the scene; e.g., sit on the armchair near the desk. HUMANISE thus enables a new generation task, language-conditioned human motion generation in 3D scenes. The proposed task is challenging as it requires joint modeling of the 3D scene, human motion, and natural language. To tackle this task, we present a novel scene-and-language conditioned generative model that can produce 3D human motions of the desirable action interacting with the specified objects. Our experiments demonstrate that our model generates diverse and semantically consistent human motions in 3D scenes.