Rajiv Mathews

Title: Innovations of Rajiv Mathews

Introduction

Rajiv Mathews is a prominent inventor based in Sunnyvale, California. He has made significant contributions to the field of automatic speech recognition (ASR) with a total of nine patents to his name. His work focuses on enhancing the accuracy and efficiency of speech recognition technologies.

Latest Patents

One of Rajiv Mathews' latest patents involves identifying and correcting ASR misrecognitions in a decentralized manner. This innovation allows on-device processors of client devices to generate predicted textual segments corresponding to spoken utterances. The system can receive further input to modify these segments and store them as candidate correction pairs. These pairs can then be transmitted to a remote system, which determines their accuracy and updates a global ASR model accordingly. Another notable patent focuses on using corrections of automated assistant functions for training on-device machine learning models. This process involves receiving sensor data, processing it to generate predicted outputs, and updating the speech recognition model based on the accuracy of these predictions.

Career Highlights

Rajiv Mathews is currently employed at Google Inc., where he continues to innovate in the field of speech recognition. His work has significantly impacted the development of more accurate and responsive automated systems.

Collaborations

Throughout his career, Rajiv has collaborated with notable colleagues, including Françoise Beaufays and Andrew Hard. These partnerships have contributed to the advancement of technologies in the speech recognition domain.

Conclusion

Rajiv Mathews is a key figure in the innovation of automatic speech recognition technologies. His patents and contributions continue to shape the future of how machines understand and process human speech.