AI Training Method Inspired by Infant Learning Boosts Performance by 15%

Penn State researchers develop a novel approach using spatial data to enhance AI's object recognition and navigation capabilities.

The new AI training method mimics how infants learn from their environment.
Models trained with this approach showed a 14.99% improvement in performance.
Researchers used virtual environments to simulate real-world learning scenarios.
The method aims to improve AI systems for exploring extreme or unfamiliar environments.
The study's datasets are available for other scientists to use in training AI models.