What is it about?

Usually artificial perception is considered a passive process of extracting information from a sensor signal. However, in robotics, the agent can interact to create and reveal its own information. Sometimes, this information first revealed through interaction is important for the success of the interaction itself, which requires an online perception loop. We present in this paper a framework to exploit interactions in online perception loops and its application to perceive kinematics and dynamics of articulated objects (doors, drawers,...) from multiple sensor modalities such as vision and haptics.

Featured Image

Why is it important?

We show that is possible to integrate interactions into perceptual systems by creating interconnected recursive subprocesses that communicate information. In realistic robot tasks, where the robot has to explore and react to the perceived information, such an online perceptual system enables manipulation with unknown objects.


This article summarizes the line of research in Interactive Perception I carried out during my PhD. The same algorithmic framework could be applied to different problems and sensor modalities. We realized the framework presents many similarities to biological perceptual systems.

Dr Roberto Martín-Martín
Stanford University

Read the Original

This page is a summary of: Coupled recursive estimation for online interactive perception of articulated objects, The International Journal of Robotics Research, May 2019, SAGE Publications,
DOI: 10.1177/0278364919848850.
You can read the full text:




The following have contributed to this page