Hamburg, Germany

Peter Doliwa

USPTO Granted Patents = 7 

Average Co-Inventor Count = 1.5

ph-index = 2

Forward Citations = 6(Granted Patents)


Company Filing History:


Years Active: 2019-2023

Loading Chart...
7 patents (USPTO):Explore Patents

Title: Innovations of Peter Doliwa

Introduction

Peter Doliwa is a notable inventor based in Hamburg, Germany. He has made significant contributions to the field of machine learning, holding a total of 7 patents. His work focuses on enhancing the understanding and functionality of machine learning models.

Latest Patents

Among his latest patents is a method for generating detailed visualizations of machine learning model behavior. This method involves inputting an image into a machine learning model for inference operations. A first heatmap is generated for the image using a visualization method, and an area of highest attention is selected based on a predetermined threshold. The selected area is then cropped and upscaled, leading to the generation of a second heatmap for analysis. Additionally, he has developed a method and machine learning system for detecting adversarial examples. This system trains two machine learning models with different data sets to identify data samples that may be adversarial in nature.

Career Highlights

Peter Doliwa is currently employed at NXP B.V., where he continues to innovate in the field of machine learning. His work has been instrumental in advancing the capabilities of machine learning systems, making them more robust and reliable.

Collaborations

He collaborates with talented individuals such as Vakul Garg and Geoffrey Thorpe, contributing to a dynamic and innovative work environment.

Conclusion

Peter Doliwa's contributions to machine learning through his patents and collaborative efforts highlight his role as a significant inventor in the field. His work continues to influence advancements in technology and machine learning applications.

This text is generated by artificial intelligence and may not be accurate.
Please report any incorrect information to support@idiyas.com
Loading…