Sotiris Malassiotis

About

Sotiris Malassiotis is a leading researcher at the intersection of computer vision and robotics, whose work has fundamentally advanced how machines perceive and manipulate highly deformable objects and cluttered environments. His primary research areas include 6D object pose estimation, robotic manipulation of textiles, and autonomous grasping in unstructured settings. Malassiotis made a landmark contribution with his work on recovering 6D object poses in crowded scenes—a paper cited over 230 times—which addressed the critical challenge of detecting objects under severe occlusion for applications in robotics and augmented reality. He is perhaps best known for pioneering the complete pipeline for autonomous clothes folding, from recognition to unfolding and folding, with multiple highly-cited papers (over 125 citations each) that introduced novel approaches using random decision forests, deep learning, and geometric methods. His innovative concept of "singulation"—isolating target objects from clutter before grasping—has been widely adopted, with his robust grasping strategies accumulating over 50 citations. Malassiotis’s work on active recognition and probabilistic planning for deformable objects, along with his geometric approaches to multifingered grasping, has established him as a key figure in enabling robots to handle the complex, real-world challenges of garment manipulation and cluttered scene interaction.

Research Focus

Key Achievements

Top Papers

1. 1
   Recovering 6D Object Pose and Predicting Next-Best-View in the Crowd

   233 citations · 2016
2. 2
   Autonomous active recognition and unfolding of clothes using random decision forests and probabilistic planning

   132 citations · 2014
3. 3
   Folding Clothes Autonomously: A Complete Pipeline

   125 citations · 2016
4. 4
   Pose and category recognition of highly deformable objects using deep learning

   64 citations · 2015
5. 5
   Fine-Grained Material Classification Using Micro-geometry and Reflectance

   60 citations · 2016
6. 6
   Robust object grasping in clutter via singulation

   53 citations · 2019
7. 7
   Active Random Forests: An Application to Autonomous Unfolding of Clothes

   45 citations · 2014
8. 8
   A geometric approach to robotic unfolding of garments

   41 citations · 2015
9. 9
   A Geometric Approach for Grasping Unknown Objects With Multifingered Hands

   40 citations · 2020
10. 10
    Total Singulation With Modular Reinforcement Learning

    25 citations · 2021

Key Collaborators

Contact & Links