Lucas Theis

This inventor holds 14 USPTO granted patents and 1 published patent application and 2 EPO patents. Top assignees: Twitter, Inc., Magic Pony Technology Limited, Google Inc.. Active years: 2020-2025.

USPTO Granted Patents = 14 (Patent Applications = 1)

% Patents Active = 35.7

EPO Patents = 2 (Patent Applications = 4)

Average Co-Inventor Count = 3.2

ph-index = 4

Forward Citations = 146(Granted Patents)

Company Filing History:

Magic Pony Technology Limited

Years Active: 2020-2025

Loading Chart...

14 patents (USPTO):Sort by Forward CitationsExplore Patents

1. 12425043 - Lossy compression with gaussian diffusion (2025-09-23),

2. 11537869 - Difference metric for machine learning-based processing systems (2022-12-27),

3. 11386599 - Feature transfer (2022-07-12),

4. 11308361 - Checkerboard artifact free sub-pixel convolution (2022-04-19),

5. 11234006 - Training end-to-end video processes (2022-01-25),

6. 11024009 - Super resolution using a generative adversarial network (2021-06-01),

7. 10909743 - Multiscale 3D texture synthesis (2021-02-02),

8. 10825138 - Super resolution using fidelity transfer (2020-11-03),

9. 10701394 - Real-time video super-resolution with spatio-temporal networks and motion compensation (2020-06-30),

10. 10699456 - Developing visual data using a hierarchical algorithm (2020-06-30),

11. 10681361 - Training end-to-end video processes (2020-06-09),

12. 10666962 - Training end-to-end video processes (2020-05-26),

13. 10623775 - End-to-end video and image compression (2020-04-14),

14. 10552977 - Fast face-morphing using neural networks (2020-02-04),

Title: The Innovative Mind of Lucas Theis

Introduction

Lucas Theis is a prominent inventor based in London, GB, known for his significant contributions to machine learning and visual data processing. With an impressive portfolio of 13 patents, his work showcases a blend of creativity and technical expertise that continues to influence the technology landscape.

Latest Patents

Among his latest patents, Lucas has developed a "Difference Metric for Machine Learning-Based Processing Systems." This invention outlines systems and methods that utilize a learned difference metric operating across a wide artifact space. The methodology involves initializing a committee of deep neural networks with labeled distortion pairs, actively learning a difference metric through psychophysics tasks, and employing this metric as an objective function in machine-learned digital file processing tasks. The process includes providing an unlabeled distortion pair to the neural networks, obtaining difference metric scores, and identifying informative distortion pairs based on diversity metrics.

Another notable patent by Lucas is titled "Feature Transfer," where he presents a method for developing visual data using source and target data coupled with a hierarchical algorithm. This method involves determining the alignment between target and source data and producing visual data by transferring features from the source to the target data after establishing this alignment. Such innovations underscore Lucas's commitment to advancing the field of machine learning and visual data processing.

Career Highlights

Lucas has held significant positions in notable companies, including Twitter, Inc. and Magic Pony Technology Limited. His work at these organizations has sharpened his skills and contributed to his expansive knowledge in the tech industry.

Collaborations

Throughout his career, Lucas has collaborated with talented individuals such as Zehan Wang and Ferenc Huszar. These partnerships have allowed him to refine his ideas and push the boundaries of innovation.

Conclusion

Lucas Theis stands out as a key figure in the field of technology and innovation. His contributions through patents and collaborative efforts signify his role as an influential inventor. As the industry continues to evolve, Lucas's work is likely to inspire future innovations in machine learning and visual data processing.