Task-Oriented Human Grasp Synthesis via Context- and Task-Aware Diffusers
This work addresses the problem of generating realistic human grasps for specific tasks, which is incremental by enhancing traditional contact maps with additional context.
The paper tackles task-oriented human grasp synthesis by introducing task-aware contact maps that incorporate scene and task information, resulting in significant improvements in grasp quality and task performance over existing methods.
In this paper, we study task-oriented human grasp synthesis, a new grasp synthesis task that demands both task and context awareness. At the core of our method is the task-aware contact maps. Unlike traditional contact maps that only reason about the manipulated object and its relation with the hand, our enhanced maps take into account scene and task information. This comprehensive map is critical for hand-object interaction, enabling accurate grasping poses that align with the task. We propose a two-stage pipeline that first constructs a task-aware contact map informed by the scene and task. In the subsequent stage, we use this contact map to synthesize task-oriented human grasps. We introduce a new dataset and a metric for the proposed task to evaluate our approach. Our experiments validate the importance of modeling both scene and task, demonstrating significant improvements over existing methods in both grasp quality and task performance. See our project page for more details: https://hcis-lab.github.io/TOHGS/