DAGDiff: Guiding Dual-Arm Grasp Diffusion to Stable and Collision-Free Grasps

2025-09-25

AI Summary

Dual-arm grasping of large or complex objects faces challenges including insufficient grasp stability, high collision risk, and poor generalization. To address these, we propose the first end-to-end SE(3)×SE(3) diffusion model specifically designed for dual-arm grasping, introducing classifier-free guidance to this task for the first time while eliminating reliance on region priors or heuristic rules. Our method jointly optimizes dual-arm grasp pose pairs during generation by integrating geometric feature extraction, force-closure stability criteria, and real-time collision awareness. Experiments demonstrate significant improvements in both simulation and real-world point clouds: +12.7% force-closure rate and +18.3% collision-free rate. The approach successfully enables a heterogeneous dual-arm system to reliably grasp and lift unseen objects, achieving physical plausibility, strong generalization, and operational safety.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Intelligent Robots: Manipulation
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Reliable dual-arm grasping is essential for manipulating large and complex objects but remains a challenging problem due to stability, collision, and generalization requirements. Prior methods typically decompose the task into two independent grasp proposals, relying on region priors or heuristics that limit generalization and provide no principled guarantee of stability. We propose DAGDiff, an end-to-end framework that directly denoises to grasp pairs in the SE(3) x SE(3) space. Our key insight is that stability and collision can be enforced more effectively by guiding the diffusion process with classifier signals, rather than relying on explicit region detection or object priors. To this end, DAGDiff integrates geometry-, stability-, and collision-aware guidance terms that steer the generative process toward grasps that are physically valid and force-closure compliant. We comprehensively evaluate DAGDiff through analytical force-closure checks, collision analysis, and large-scale physics-based simulations, showing consistent improvements over previous work on these metrics. Finally, we demonstrate that our framework generates dual-arm grasps directly on real-world point clouds of previously unseen objects, which are executed on a heterogeneous dual-arm setup where two manipulators reliably grasp and lift them.

Problem

• Achieving stable dual-arm grasping for large complex objects
• Preventing collisions between arms during coordinated grasping
• Generalizing grasp generation to unseen objects without priors

Innovation

• Denoises grasp pairs in SE(3) x SE(3) space
• Guides diffusion with classifier signals for stability
• Integrates geometry, stability, and collision-aware guidance