Arul Selvam Periyasamy

About

Arul Selvam Periyasamy is a robotics researcher whose work sits at the intersection of computer vision, deep learning, and autonomous robotic manipulation. His research focuses primarily on object detection, semantic segmentation, and 6D pose estimation — the critical perceptual capabilities that enable robots to identify, locate, and interact with objects in complex, cluttered environments. Periyasamy's most influential contribution, "RGB-D Object Detection and Semantic Segmentation for Autonomous Manipulation in Clutter" (2017, 165 citations), established deep learning pipelines for robots navigating visually complex scenes with partial occlusions. This work laid important groundwork for practical bin-picking applications, further demonstrated through his team's acclaimed entry in the 2017 Amazon Robotics Challenge (82 citations), which showcased dual-arm coordination and rapid object learning in warehouse-like settings. His more recent research has advanced transformer-based architectures for multi-object 6D pose estimation, reflected in the YOLOPose series (2023), and he has tackled the nuanced challenge of symmetric object orientation estimation without pose labels. A member of the renowned NimbRo robotics team at the University of Bonn, Periyasamy contributed to award-winning entries at both the DARPA Robotics Challenge and MBZIRC 2017, underscoring his impact on competitive, real-world autonomous robotics.

Research Focus

Key Achievements

Top Papers

1. 1
   RGB-D object detection and semantic segmentation for autonomous manipulation in clutter

   165 citations · 2017
2. 2
   Fast Object Learning and Dual-arm Coordination for Cluttered Stowing, Picking, and Packing

   82 citations · 2018
3. 3
   Supervised Autonomy for Exploration and Mobile Manipulation in Rough Terrain with a Centaur-Like Robot

   30 citations · 2016
4. 4
   YOLOPose: Transformer-Based Multi-object 6D Pose Estimation Using Keypoint Regression

   28 citations · 2023
5. 5
   YOLOPose V2: Understanding and improving transformer-based 6D pose estimation

   28 citations · 2023
6. 6
   Robust 6D Object Pose Estimation in Cluttered Scenes Using Semantic Segmentation and Pose Regression Networks

   22 citations · 2018
7. 7
   Team NimbRo at MBZIRC 2017: Autonomous valve stem turning using a wrench

   16 citations · 2018
8. 8
   Learning Implicit Probability Distribution Functions for Symmetric Orientation Estimation from RGB Images Without Pose Labels

   3 citations · 2022
9. 9
   T6D-Direct: Transformers for Multi-object 6D Pose Direct Regression

   3 citations · 2021
10. 10
    ConvPoseCNN2: Prediction and Refinement of Dense 6D Object Pose

    3 citations · 2022

Key Collaborators

Contact & Links