GCNGrasp-VP: Affordance-Guided View Planning for Efficient Task-Oriented Grasping

2026-06-17

AI Summary

Task-oriented grasping suffers significant performance degradation when key object regions are occluded. Existing approaches either rely on assumptions of initial visibility or require time-consuming scene reconstruction while neglecting task semantics. This work proposes GCNGrasp-VP, a novel framework that introduces utility field prediction into viewpoint planning for the first time. By integrating the GCNGrasp-v2 grasping model with a lightweight information gain metric, the method achieves constant-time inference without explicit scene reconstruction. Requiring only a single viewpoint adjustment, it efficiently observes task-relevant regions and substantially outperforms uncertainty-based baselines in single-object scenarios, markedly improving grasp success rates while maintaining millisecond-level latency.

Abstract

Task-oriented grasping performance degrades significantly when object views suffer from occlusions. Existing task-oriented grasping methods typically assume task-relevant regions are visible in the initial frame, while view planning approaches enable active perception but often ignore task semantics and rely on time-consuming scene reconstruction. To address these limitations, we present GCNGrasp-VP, an efficient framework integrating affordance field prediction with active view planning. Central to this framework is GCNGrasp-v2, a task-oriented grasp model that simultaneously supports grasp evaluation and affordance field prediction, achieving constant-time inference complexity. Leveraging this capability, our Affordance-guided View Planner (Affordance-VP) utilizes the affordance field as an information gain metric to guide camera observation of task-relevant regions without requiring scene reconstruction. View planning results show that our method significantly outperforms scene-uncertainty-driven baselines with only one view adjustment. Real-world validation further confirms substantial improvements in grasp success rates for single-object scenarios while maintaining millisecond-level computational latency. Code and models are available at https://github.com/Instinct323/GCNGrasp-VP.

Problem

• task-oriented grasping
• occlusion
• view planning
• affordance
• active perception

Innovation

• affordance-guided view planning
• task-oriented grasping
• constant-time inference
• active perception
• grasp affordance field