Michael Smart
Papers
1
Total Citations
2
H-Index
1
About
Michael Smart is an emerging researcher working at the intersection of deep learning, robotics, and probabilistic reasoning. His work focuses on uncertainty estimation in neural network systems, particularly as applied to autonomous and robotic platforms. Smart's most notable contribution is "BayesOD: A Bayesian Approach for Uncertainty Estimation in Deep Object Detectors" (2019), which addresses a critical challenge in deploying deep neural networks within real-world robotic systems: the absence of reliable uncertainty measures associated with model predictions. By introducing a Bayesian framework tailored to deep object detection pipelines, Smart's research pushes toward safer and more trustworthy AI-driven systems — a problem of growing importance as autonomous robots and vehicles become more prevalent. While still in the early stages of accumulating citations, his work tackles a foundational limitation in the field, bridging the gap between high-performing but overconfident neural networks and the reliability demands of safety-critical applications. Smart's research represents a meaningful step forward in making deep learning systems not only accurate, but appropriately calibrated and self-aware about the limits of their own predictions.
Research Focus
Key Achievements
Top Papers
- 1