Aleksandar Vakanski

About

Aleksandar Vakanski is a leading researcher in the field of **robot programming by demonstration (PbD)** and **robotic learning from visual observation**. His work focuses on enabling robots to acquire complex manipulation skills by observing human demonstrations, rather than through explicit manual programming. Vakanski’s major contributions include pioneering the use of **Hidden Markov Models (HMM)** and **Dynamic Time Warping (DTW)** for learning and reproducing complex trajectories, a method detailed in his highly influential 2012 paper, which has garnered **158 citations**. He has also advanced the application of **Conditional Random Fields (CRF)** for transferring motor skills to robots, and developed image-based trajectory planning approaches for robust PbD. Notably, his research integrates **visual perception** using sensors like the Microsoft Kinect to track object manipulation, and addresses the challenging problem of **6D object pose estimation** for robotic grasping. With a total of over **210 citations** across his most-cited works, Vakanski’s research is foundational for creating more intuitive, flexible, and autonomous robotic systems, directly impacting the future of manufacturing, healthcare, and service robotics.

Research Focus

Key Achievements

Top Papers

1. 1
   Trajectory Learning for Robot Programming by Demonstration Using Hidden Markov Model and Dynamic Time Warping

   158 citations · 2012
2. 2
   An image-based trajectory planning approach for robust robot programming by demonstration

   21 citations · 2017
3. 3
   Robot Learning by Visual Observation

   12 citations · 2017
4. 4
   Trajectory Learning based on Conditional Random Fields for Robot Programming by Demonstration

   12 citations · 2010
5. 5
   Robotic Learning of Manipulation Tasks from Visual Perception Using a Kinect Sensor

   5 citations · 2014
6. 6
   6D Object Pose Estimation for Robot Programming by Demonstration

   2 citations · 2019

Key Collaborators

Contact & Links