Matt Unrath
Papers
1
Total Citations
3
H-Index
1
About
Matt Unrath is a researcher at the intersection of robotics, machine learning, and human-computer interaction, with a primary focus on scalable data generation for autonomous systems. His most cited work, "Using Crowdsourcing to Generate Surrogate Training Data for Robotic Grasp Prediction" (2014, 3 citations), introduces a novel approach to overcoming a critical bottleneck in robotic learning: the expensive and time-consuming collection of physical training data. By leveraging crowd-sourced evaluations of grasp images, Unrath demonstrates that surrogate data can effectively train models for grasp quality prediction, particularly in constrained regions of the grasp space. This contribution highlights his broader interest in democratizing robotics research through human-in-the-loop systems. While his citation count reflects a niche but foundational contribution, Unrath’s work is notable for its early exploration of crowdsourcing as a practical alternative to real-world robotic experimentation—a concept that has since gained traction in the field. His research underscores the potential of combining human intelligence with machine learning to accelerate progress in robotic manipulation.
Research Focus
Key Achievements
Top Papers
- 1