Helen Amelia Hawkins

Title: Helen Amelia Hawkins: Inventor in Machine-Learning Training Data

Introduction

Helen Amelia Hawkins is an inventor associated with Huntsville, Alabama, and The Boeing Company. Her patent work addresses the generation of training data for machine-learning systems, including data representing unusual or edge-case scenarios.

Latest Patents

Hawkins is associated with 1 patent: “Computing Device, Method and Computer Program Product for Generating Training Data for a Machine Learning System.” The invention describes a computing device that uses a simulator and a sampling algorithm to create varied scenes within a scenario defined by parametric attributes. A physics engine generates training data for those scenes, including edge scenes, and modifies the attributes to produce additional data from further simulated scenarios.

Career Highlights

Hawkins’s listed patent reflects work at the intersection of simulation, physics-based modeling, and machine learning. The approach is designed to help generate diverse training examples, including examples of less common conditions that may be important for training a machine-learning system.

Collaborations

Hawkins is associated with The Boeing Company. Her listed coworkers include Patrick Daniel Dees and Taylor S. Loper.

Conclusion

Helen Amelia Hawkins’s patent highlights a simulation-based method for generating varied machine-learning training data, including data representing edge scenes.