VLM-integrated 3D perception model for robust robotic grasping adapted to deformable sacks with arbitrary shapes

Key Points

• Robust robotic grasping is achieved through a visual language model integrated with 3D perception, enhancing operational efficiency.
• The model demonstrates efficacy in manipulating deformable sacks with various shapes, crucial for versatile applications.
• Analysis incorporates diverse scenarios to validate experimental outcomes, leading to improved accuracy and adaptability.
• Future work may enable broader applications in logistics and automation, supporting evolving demands.