This inventor holds 1 USPTO granted patent. Top assignee: Robert Bosch. Active years: 2026.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2026
Title: Reiko Lettmoden: Inventor in Neural Network Training
Introduction
Reiko Lettmoden is an inventor associated with Braunschweig, Germany, and Robert Bosch. Her patent work addresses methods for training machine-learning systems to transfer images between domains.
Latest Patents
Lettmoden is associated with 1 patent, titled “Method and Device for Training a Neural Network.” The invention describes a computer-implemented method for training a machine-learning system to transfer images from a source domain into a target domain.
The method identifies patches from source and target images, with each patch assigned pixel-by-pixel metadata. It then forms tuples pairing a source patch with one or more target patches selected as its nearest neighbors according to a similarity measure based on that metadata. The system is trained using the source and target patches in those tuples.
Career Highlights
Lettmoden’s listed patent reflects work on image-domain transfer and neural-network training. The disclosed approach uses metadata-based patch similarity to guide the selection of training examples.
Collaborations
Lettmoden is associated with Robert Bosch. Her listed coworkers include Maximilian Menke and Thomas Wenzel.
Conclusion
Reiko Lettmoden’s patent highlights a patch-based approach to training neural networks for transferring images between domains.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile