Karthik Mahesh Varadarajan

About

Karthik Mahesh Varadarajan is a leading researcher in cognitive robotics and affordance-based perception, whose work bridges the gap between human-like understanding and machine autonomy. His primary research areas include affordance theory, robotic manipulation, 3D scene understanding, and human-robot interaction. Varadarajan’s most significant contribution is the creation of **AfNet (The Affordance Network)**, an open-source framework that defines over 250 household objects in terms of 35 grounded affordance features—or “afbits”—enabling robots to recognize objects by their functional possibilities rather than just appearance. This foundational work, detailed in papers like *“AfRob: The Affordance Network Ontology for Robots”* (46 citations) and *“AfNet: The Affordance Network”* (31 citations), has become a key resource for cognitive visual recognition systems. He also advanced grasp planning through affordance-based part recognition and topological mapping for robot navigation. Beyond affordances, Varajan has contributed to augmented virtuality for mining robot telepresence and real-time depth diffusion for 3D surface reconstruction. His work on the Cognitive Object Recognition System (CORS) integrates multiple recognition algorithms inspired by neurocomputational models. With over 140 total citations across his top publications, Varadarajan’s research continues to shape how robots perceive, reason about, and interact with their environments.

Research Focus

Key Achievements

Top Papers

1. 1
   AfRob: The affordance network ontology for robots

   46 citations · 2012
2. 2
   AfNet: The Affordance Network

   31 citations · 2013
3. 3
   Affordance based Part Recognition for Grasping and Manipulation

   13 citations · 2011
4. 4
   Topological mapping for robot navigation using affordance features

   13 citations · 2015
5. 5
   Knowledge Representation and Inference for Grasp Affordances

   9 citations · 2011
6. 6
   Augmented virtuality based immersive telepresence for control of mining robots

   9 citations · 2011
7. 7
   Combining Plane Estimation with Shape Detection for Holistic Scene Understanding

   8 citations · 2011
8. 8
   Real-time depth diffusion for 3D surface reconstruction

   8 citations · 2010
9. 9
   Cognitive object recognition system (CORS)

   6 citations · 2010
10. 10
    Attention driven grasping for clearing a heap of objects

    4 citations · 2012

Key Collaborators

Contact & Links