Zurich, Switzerland

Mikhail Dektiarev

This inventor holds 1 USPTO granted patent. Top assignee: Google Inc.. Active years: 2026.


% Patents Active = 100.0

Average Co-Inventor Count = 2.0

ph-index = 1


Company Filing History:


Years Active: 2026

Loading Chart...
1 patent (USPTO):Explore Patents

Title: Mikhail Dektiarev: Innovator in Speech Recognition Technology

Introduction

Mikhail Dektiarev is a prominent inventor based in Zurich, Switzerland. He has made significant contributions to the field of speech recognition technology. His innovative approach focuses on enhancing the accuracy of speech recognition systems through the use of voice query history.

Latest Patents

Dektiarev holds a patent for a method titled "Voice-history based speech biasing." This method improves speech recognition by utilizing voice query history. It involves receiving audio data from a current query spoken by a user and processing this data to create a lattice of candidate hypotheses. The method also incorporates voice query history data, which includes n-grams extracted from previous queries. By generating a biasing context vector, the system can indicate the likelihood of certain n-grams appearing in the current query. This approach augments the lattice of candidate hypotheses, ultimately leading to a more accurate transcription of the user's current query.

Career Highlights

Mikhail Dektiarev is currently employed at Google Inc., where he continues to develop innovative solutions in the realm of speech recognition. His work has garnered attention for its potential to significantly enhance user experience in voice-activated technologies.

Collaborations

Dektiarev collaborates with Ágoston Weisz, a fellow innovator in the field. Their partnership exemplifies the collaborative spirit that drives advancements in technology.

Conclusion

Mikhail Dektiarev's contributions to speech recognition technology highlight the importance of innovation in improving user interactions with technology. His work continues to pave the way for more intuitive and effective speech recognition systems.

Profile summary based on public USPTO records.
Please report any incorrect information to [email protected]
Loading…