Company Filing History:
Years Active: 2024-2025
Title: **Ehsan Amid: Pioneering Innovations in Automatic Speech Recognition**
Introduction
Ehsan Amid is a distinguished inventor based in Mountain View, CA, gaining recognition for his impactful contributions to the field of automatic speech recognition (ASR). With two patents under his name, Amid has advanced techniques that enhance the performance of ASR models, showcasing his dedication to innovation and technology.
Latest Patents
Ehsan Amid's latest innovation revolves around a method for phrase extraction in ASR models. This sophisticated method involves obtaining audio data that characterizes an utterance and its corresponding ground-truth transcription. The audio data is modified to obfuscate a particular phrase recited during the utterance. Utilizing a trained ASR model, the modified audio data is processed to generate a predicted transcription of the utterance. A critical aspect of this method is determining if the predicted transcription contains the particular phrase by comparing it to the ground-truth transcription. If the predicted transcription does include the phrase, the process generates an output indicating that the trained ASR model has leaked the particular phrase from the training dataset. This innovative approach significantly enhances the privacy and security features of ASR systems.
Career Highlights
Ehsan Amid works at Google Inc., a leading technology company where he applies his expertise in ASR and machine learning. His role involves continual development and refinement of algorithms that power speech recognition technologies, facilitating a more seamless interaction between humans and machines.
Collaborations
Throughout his career, Ehsan has collaborated with notable colleagues, including Rajiv Mathews and Francoise Beaufays. These partnerships within Google Inc. underscore the collective effort in advancing ASR technologies and contribute to a dynamic innovation environment.
Conclusion
Ehsan Amid stands out as a promising inventor in the realm of automatic speech recognition, with his innovative patents poised to influence the future of human-computer interaction. His contributions not only demonstrate technical proficiency but also reflect a commitment to enhancing the functionality and security of speech recognition systems.