Connor Schenck

About

Connor Schenck is a robotics researcher whose work spans robot perception, multimodal learning, and manipulation of complex materials. He is perhaps best known for his pioneering contributions to interactive object recognition, where he developed frameworks enabling robots to identify and categorize household objects through behavioral exploration — combining proprioceptive, auditory, and visual feedback rather than relying on passive sensing alone. His highly cited 2012 study grounding semantic categories across 100 objects (74 citations) and his 2011 work on proprioceptive and auditory-based recognition (63 citations) established him as a leading voice in embodied, interaction-driven robot learning. Schenck extended this work to relational object categories, teaching robots to understand not just individual objects but pairwise and group-level relationships through multimodal perception (55 citations). A distinctive thread in his later research is liquid and granular media perception — an underexplored but practically vital domain. Applying fully convolutional deep neural networks, he developed methods for detecting, tracking, and reasoning about liquids, and tackled the challenging problem of pouring control using visual feedback. His work on granular media manipulation further demonstrates his commitment to real-world robustness. Across his career, Schenck has consistently pushed robots toward richer, more human-like understanding of their physical environments.

Research Focus

Key Achievements

Top Papers

1. 1
   Grounding semantic categories in behavioral interactions: Experiments with 100 objects

   74 citations · 2012
2. 2
   Interactive object recognition using proprioceptive and auditory feedback

   63 citations · 2011
3. 3
   Learning relational object categories using behavioral exploration and multimodal perception

   55 citations · 2014
4. 4
   Perceiving and reasoning about liquids using fully convolutional networks

   30 citations · 2017
5. 5
   Learning Robotic Manipulation of Granular Media

   24 citations · 2017
6. 6
   Detection and Tracking of Liquids with Fully Convolutional Networks

   14 citations · 2016
7. 7
   Which Object Fits Best? Solving Matrix Completion Tasks with a Humanoid Robot

   12 citations · 2014
8. 8
   Which Object Comes Next? Grounded Order Completion by a Humanoid Robot

   6 citations · 2012
9. 9
   Visual closed-loop control for pouring liquids

   6 citations · 2017
10. 10
    Intelligence tests for robots: Solving perceptual reasoning tasks with a humanoid robot

    5 citations · 2013

Key Collaborators

Contact & Links