Over the past five years, significant progress has been made in deformable object (DO) manipulation, spanning breakthroughs in representation, simulation, and control, as well as the integration of machine learning and large models for improved perception. Despite these advancements, key challenges persist, particularly in developing a unified framework for DO representation across different domains.
The “5th Workshop: Reflections on Representations and Manipulating Deformable Objects” aims to critically review recent developments and identify future research priorities. Key topics will include state representation, non-linear dynamic modeling, and the application of advanced models like vision-language models (VLMs).
The workshop will highlight progress in learning-based methods, such as reinforcement and imitation learning, which have enhanced robotic dexterity in handling deformable materials. Additionally, it will explore the role of low-cost hardware and open datasets in accelerating research and making advancements more accessible.
By bringing together researchers and industry experts, this workshop seeks to shape the future of DO manipulation, fostering collaboration and innovation in this rapidly evolving field.
