SIS: Seam-Informed Strategy for T-shirt Unfolding

2024-09-11

IEEE Robotics and Automation Letters

AI Summary

Automated unfolding of deformable garments such as T-shirts remains challenging due to the difficulty of selecting effective grasping points. Method: This paper proposes a dual-arm collaborative manipulation framework leveraging seam geometry and topology as explicit, task-driven structural cues. We introduce a Seam Feature Extraction Module (SFEM) to model seams explicitly and design a Decision Matrix Initialization and Iterative Update Mechanism (DMIM), initialized via human demonstrations and refined through execution feedback, enabling real-world data-driven grasp policy learning. Crucially, the approach integrates coordinated bimanual motion planning without relying on simulation-based pretraining. Contribution/Results: Experiments on a physical robot platform demonstrate a significant improvement in T-shirt unfolding success rate over baseline methods lacking seam priors, empirically validating seams as robust, semantically meaningful cues for garment manipulation.

Technology Category

• Intelligent Robots: Manipulation
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Reasoning under Uncertainty: Sequential Decision Making

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Seams are information-rich components of garments. The presence of different types of seams and their combinations helps to select grasping points for garment handling. In this paper, we propose a new Seam-Informed Strategy (SIS) for finding actions for handling a garment, such as grasping and unfolding a T-shirt. Candidates for a pair of grasping points for a dual-arm manipulator system are extracted using the proposed Seam Feature Extraction Method (SFEM). A pair of grasping points for the robot system is selected by the proposed Decision Matrix Iteration Method (DMIM). The decision matrix is first computed by multiple human demonstrations and updated by the robot execution results to improve the grasping and unfolding performance of the robot. Note that the proposed scheme is trained on real data without relying on simulation. Experimental results demonstrate the effectiveness of the proposed strategy. The project video is available at https://github.com/lancexz/sis.

Problem

• Develops a seam-informed strategy for T-shirt unfolding
• Extracts grasping points using Seam Feature Extraction Method
• Improves robot performance via human demonstrations and execution updates

Innovation

• Uses Seam Feature Extraction Method (SFEM)
• Applies Decision Matrix Iteration Method (DMIM)
• Trains on real data without simulation