Somerville, MA, United States of America

Andrew Jake Rosenbaum

This inventor holds 1 USPTO granted patent. Top assignee: Amazon Technologies, Inc.. Active years: 2016.


Average Co-Inventor Count = 4.0

ph-index = 1

Forward Citations = 57(Granted Patents)


Company Filing History:


Years Active: 2016

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

Title: Andrew Jake Rosenbaum: Innovator in Active Learning for Speech Recognition

Introduction

Andrew Jake Rosenbaum is an accomplished inventor based in Somerville, MA (US). He has made significant contributions to the field of automatic speech recognition (ASR) through his innovative patent. His work focuses on enhancing the efficiency of speech recognition systems, particularly in the area of lexical annotations.

Latest Patents

Rosenbaum holds a patent titled "Active learning for lexical annotations." This patent discloses features for active learning that identify words likely to improve guessing and ASR after manual annotation. In traditional speech recognition systems, a lexicon is used to provide pronunciations for known words. For unknown words, pronunciation-guessing (G2P) methods may be employed to generate pronunciations automatically. However, manually annotated pronunciations yield better ASR results than automatic ones, which can sometimes be inaccurate. The active learning features included in his patent help direct limited annotation resources effectively.

Career Highlights

Rosenbaum is currently employed at Amazon Technologies, Inc., where he continues to develop innovative solutions in the realm of speech recognition. His expertise in active learning and lexical annotations has positioned him as a valuable asset in the technology sector.

Collaborations

Throughout his career, Rosenbaum has collaborated with notable colleagues, including Alok Ulhas Parlikar and Jeffrey Paul Lilly. These collaborations have further enriched his work and contributed to advancements in the field.

Conclusion

Andrew Jake Rosenbaum is a pioneering inventor whose work in active learning for lexical annotations has the potential to significantly enhance automatic speech recognition systems. His contributions continue to shape the future of speech technology.

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