Kevin Galassi
Papers
14
Total Citations
147
H-Index
5
About
Kevin Galassi is a robotics researcher whose work centers on the perception and manipulation of Deformable Linear Objects (DLOs) — a challenging and underexplored domain encompassing cables, wires, ropes, and elastic tubes ubiquitous in industrial and domestic settings. His research addresses the fundamental difficulty robotic systems face in handling such objects, combining computer vision, machine learning, and model-based control to enable autonomous manipulation in real-world scenarios. Among his most impactful contributions is RT-DLO, a real-time instance segmentation framework that has garnered 33 citations since 2023, alongside a manipulation framework featuring online model parameter estimation (32 citations), demonstrating his ability to bridge perception and control. His earlier work on robotic switchgear cabling and wiring harness manufacturing (25 citations) highlighted the industrial relevance of his research, while his DLO3DS approach brought multi-view 3D shape tracking to practical deployment. More recently, Galassi has expanded his scope to include cloth manipulation, large language model grounding for DLO perception, and automated error detection in assembly pipelines. With over 130 cumulative citations and a growing portfolio spanning dual-arm robotics and scalable learning architectures, Galassi is emerging as a significant voice in the automation of deformable object manipulation.
Research Focus
Key Achievements
Top Papers
- 1RT-DLO: Real-Time Deformable Linear Objects Instance Segmentation33 citations · 2023
- 2
- 3
- 4
- 53D DLO Shape Detection and Grasp Planning from Multiple 2D Views12 citations · 2021
- 6Cable Detection and Manipulation for DLO-in-Hole Assembly Tasks4 citations · 2022
- 7
- 8
- 9
- 10