Aaron Bobick

About

Aaron Bobick is a prominent researcher in robotics, computer vision, and autonomous systems, whose work spans more than three decades of influential contributions to the field. His research focuses on several interconnected domains: autonomous robot navigation, human-robot collaboration, activity recognition, and affordance learning. Bobick's early work on representation paradigms for concurrent object descriptions (1992) laid conceptual groundwork for how robots perceive and model their environments dynamically. He later made significant strides in outdoor robot autonomy, developing unsupervised visual learning methods for terrain traversability classification (160 citations), enabling robots to navigate unstructured environments more reliably. A particularly rich thread of his research involves anticipating and modeling human behavior in collaborative settings—his probabilistic frameworks for predicting human actions and wait-sensitive planning (87 citations) have proven highly influential in responsive human-robot interaction. His contributions to activity parsing using stochastic grammars and Bayes networks further demonstrate a sophisticated approach to understanding complex, multi-step human behaviors. Bobick also advanced robot manipulation through work on object singulation, contact-location learning, and affordance prediction from visual categories. His involvement in the DARPA Robotics Challenge underscores his commitment to translating research into real-world humanoid robot applications.

Research Focus

Key Achievements

Top Papers

1. 1
   Traversability classification using unsupervised on-line visual learning for outdoor robot navigation

   160 citations · 2006
2. 2
   Probabilistic human action prediction and wait-sensitive planning for responsive human-robot collaboration

   87 citations · 2013
3. 3
   From Stochastic Grammar to Bayes Network: Probabilistic Parsing of Complex Activity

   82 citations · 2014
4. 4
   Learning Visual Object Categories for Robot Affordance Prediction

   79 citations · 2009
5. 5
   Anticipating human actions for collaboration in the presence of task and sensor uncertainty

   78 citations · 2014
6. 6
   A General‐purpose System for Teleoperation of the DRC‐HUBO Humanoid Robot

   77 citations · 2015
7. 7
   Guided pushing for object singulation

   57 citations · 2012
8. 8
   Learning contact locations for pushing and orienting unknown objects

   45 citations · 2013
9. 9
   The representation space paradigm of concurrent evolving object descriptions

   42 citations · 1992
10. 10
    Decoupling behavior, perception, and control for autonomous learning of affordances

    30 citations · 2013

Key Collaborators

Contact & Links