Ervin Teng

Papers

Total Citations

H-Index

About

Ervin Teng is a researcher working at the intersection of robotics, machine learning, and autonomous systems, with a particular focus on enabling intelligent behavior in resource-constrained robotic platforms. His work explores how robots can develop curiosity-driven learning capabilities, allowing them to autonomously seek out informative experiences rather than passively relying on human-provided training data. His 2019 paper, "Autonomous Curiosity for Real-Time Training Onboard Robotic Agents," addresses a fundamental challenge in applied machine learning: reducing the dependency on human operators to supply ground truth labels by equipping robots with the ability to identify and pursue the most valuable new information themselves. This contribution is especially significant in real-world deployment scenarios where human supervision is costly or impractical. While still an emerging body of work with early citation counts, Teng's research tackles a genuinely difficult problem — balancing computational efficiency with adaptive learning on embedded hardware. His contributions lay important groundwork for the development of more self-sufficient robotic systems capable of continuous, on-device learning, a capability increasingly vital as autonomous robots are deployed in dynamic and unpredictable environments.