Pramod Chandrashekhariah

About

Pramod Chandrashekhariah is a researcher at the intersection of developmental robotics, active vision, and autonomous learning. His work is fundamentally concerned with endowing robots with the intrinsic curiosity and learning mechanisms of human infants, enabling them to autonomously explore and understand their visual world. A major contribution is his development of a "curious" active vision system for humanoid robots, which learns object representations without human assistance by using an attention mechanism driven by learning progress. This concept, formalized in his theory of the "familiarity-to-novelty shift," explains how infants’ interest in stimuli is governed by the improvement of their internal models. His most cited work (24 citations) introduces a method for autonomously learning active multi-scale binocular vision, allowing a robot to learn visual disparity and vergence movements to fixate on objects. Further contributions include robust visual object detection and real-time activity recognition using deep learning of motion features. Chandrashekhariah’s research uniquely bridges computational modeling and developmental psychology, creating systems that learn as naturally as a child.

Research Focus

Key Achievements

Top Papers

1. 1
   Autonomous learning of active multi-scale binocular vision

   24 citations · 2013
2. 2
   Autonomous learning of robust visual object detection and identification on a humanoid

   10 citations · 2012
3. 3
   Learning visual object detection and localisation using icVision

   7 citations · 2013
4. 4
   Familiarity-to-novelty shift driven by learning: A conceptual and computational model

   5 citations · 2011
5. 5
   Let it Learn - A Curious Vision System for Autonomous Object Learning

   3 citations · 2013
6. 6
   Real-time activity recognition via deep learning of motion features

   3 citations · 2015

Key Collaborators

Contact & Links