Seg2Grasp: A Robust Modular Suction Grasping in Bin Picking
2026-07-20 • Robotics
RoboticsArtificial Intelligence
AI summaryⓘ
The authors created Seg2Grasp, a step-by-step method to help robots pick up objects using suction in messy bins. First, their system finds objects clearly in pictures using a Transformer model. Then, it chooses the best spots to grab the objects based on their shapes. Lastly, it identifies what the objects are using a special classifier called Mask-CLIP. Tests show their approach works better than others for different and complicated items in real-world settings.
bin pickingsuction graspingTransformer modelRGB-D imagessurface normalsobject segmentationMask-CLIPopen-vocabulary classificationrobotic graspingmodular pipeline
Authors
Hye-Jung Yoon, Juno Kim, Yesol Park, Jun-Ki Lee, Byoung-Tak Zhang
Abstract
Current bin picking methods that rely heavily on end-to-end learning often falter when confronted with unfamiliar or complex objects in unstructured environments. To overcome these limitations, we introduce Seg2Grasp, a modular pipeline designed for robust suction grasping in dynamic and cluttered bin scenarios. Seg2Grasp is built on a three-step process: Segmentation, Grasping, and Classification. The Segmentation module employs a Transformer-based model to generate class-agnostic object masks from RGB-D images, ensuring accurate detection across various conditions. The Grasping module uses surface normals and mask proposals to determine the optimal suction points, enhancing grasp success. Finally, the Classification module leverages fine-tuned open-vocabulary Mask-CLIP for precise object identification, enabling versatile handling of diverse objects. Real-world robotic experiments demonstrate that Seg2Grasp outperforms existing methods in success rates and adaptability, establishing it as a powerful tool for automated bin picking in industrial settings.