Behnam Asefisaray

Title: Behnam Asefisaray: Innovator in Automatic Speech Recognition Technology

Introduction

Behnam Asefisaray is a notable inventor based in Cambridge, GB. He has made significant contributions to the field of automatic speech recognition (ASR) technology. His innovative work focuses on improving the accuracy of ASR outputs through advanced methodologies.

Latest Patents

Asefisaray holds a patent for a "Transcription Knowledge Graph." This invention addresses the challenges associated with ASR engine output errors. The transcription knowledge graph utilizes data from previous ASR sessions to create a voice graph. This graph can analyze the correlation between mis-transcriptions and correct transcriptions. Consequently, even if the ASR output contains errors, it can be adjusted to reflect the correct transcription. Additionally, the correct transcriptions and voice graph can be employed to train machine learning algorithms. These algorithms can generate numerical representations of entities, enhancing the identification of corresponding entity labels.

Career Highlights

Behnam Asefisaray is currently employed at Roku, Inc., where he continues to develop innovative solutions in the realm of speech recognition technology. His work has the potential to significantly enhance user experiences in various applications that rely on ASR systems.

Collaborations

Asefisaray collaborates with talented individuals such as Doo Soon Kim and Minsuk Heo. Their combined expertise contributes to the advancement of technologies that improve speech recognition accuracy.

Conclusion

Behnam Asefisaray is a pioneering inventor whose work in automatic speech recognition technology is making a substantial impact. His innovative patent for a transcription knowledge graph exemplifies his commitment to enhancing the accuracy of ASR systems.