Meilen, Switzerland

Dominik Roblek

This inventor holds 78 USPTO granted patents and 6 published patent applications and 5 EPO patents, primarily in Audio Processing (CPC class G06F). Top assignees: Google Inc., Waymo LLC, Other. Active years: 2014-2025.

USPTO Granted Patents = 78 

% Patents Active = 92.3

 

Average Co-Inventor Count = 3.0

ph-index = 12

Forward Citations = 845(Granted Patents)

Forward Citations (Not Self Cited) = 762(Dec 10, 2025)


Inventors with similar research interests:


Location History:

  • Zürich, CH (2014)
  • Kilchberg ZH, CH (2015 - 2016)
  • Mountain View, CA (US) (2014 - 2018)
  • Ruschlikon, CH (2014 - 2020)
  • Meilen, CH (2017 - 2024)
  • Kilchberg, CH (2019 - 2024)

Company Filing History:


Years Active: 2014-2025

Loading Chart...
Loading Chart...
Areas of Expertise:
Audio Processing
Neural Networks
Speaker Verification
Keyword Spotters
Audio Waveforms
Music Identification
Self-Supervised Learning
Data Compression
Entity Repository
Content Aggregation
Audio Classification
Machine Learning
78 patents (USPTO):Explore Patents

Title: Dominik Roblek: A Pioneer in Neural Network Innovations

Introduction

Dominik Roblek, based in Meilen, Switzerland, has established himself as a prominent inventor with an impressive portfolio of 77 patents. His groundbreaking work primarily focuses on innovations in neural networks and their applications in data processing and audio generation.

Latest Patents

Among his latest patents, Dominik has developed methods, systems, and apparatus for generating coded data representations using neural networks and vector quantizers. This innovative approach involves receiving a new input, processing it with an encoder neural network to produce a feature vector, and then generating a coded representation through a series of vector quantizers linked to respective codebooks. This method effectively defines a quantized representation of the feature vector.

In another significant patent, he created methods and systems for generating audio waveforms utilizing encoder and decoder neural networks. This technology processes an input audio waveform through a generator neural network, culminating in the production of an output audio waveform. The process involves receiving an input audio waveform, generating feature vectors, and using a decoder neural network to achieve the final output, ensuring each output time step corresponds to an audio sample.

Career Highlights

Dominik has made noteworthy contributions during his career, notably at Google Inc. and Waymo LLC, where he applied his expertise in neural networks and machine learning to advance cutting-edge technologies. His work at these prestigious companies has solidified his reputation as a leader in innovation.

Collaborations

Throughout his career, Dominik has collaborated with talented professionals including Matthew Nirvan Sharifi and Annie Chen. These partnerships have fostered an environment of creativity and innovation, allowing for the development of transformative technologies in the field of artificial intelligence and beyond.

Conclusion

Dominik Roblek's contributions to the field of neural networks and his extensive patent portfolio underscore his role as an influential inventor. His innovative approaches to generating coded data representations and audio waveforms mark significant advancements in technology, promising to shape the future of machine learning and artificial intelligence.

This text is generated by artificial intelligence and may not be accurate.
Please report any incorrect information to [email protected]
Loading…