New York, NY, United States of America

Hainan Xu

This inventor holds 1 USPTO granted patent. Top assignee: Nvidia Corporation. Active years: 2026.

IDiyas Innovation Intelligence. (2026). Inventor Profile: Hainan Xu. Retrieved from https://idiyas.com/inventor/hainan-xu-bq8soo90

Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile

Calculating rank

Average Co-Inventor Count = 1.0


Company Filing History:


Years Active: 2026

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

Title: Hainan Xu and Pronunciation-Aware Conversational AI

Introduction

Hainan Xu is an inventor associated with New York, NY, and Nvidia Corporation. His patent work addresses how conversational AI systems can use pronunciation information to improve speech recognition.

Latest Patents

Xu is associated with 1 patent: “Pronunciation-Aware Embedding Generation for Conversational AI Systems and Applications.” The invention describes receiving an acoustic frame and prior textual tokens, then generating a multidimensional embedding that captures characteristics of those tokens, including pronunciation information. A system can use that embedding alongside an encoded representation of the current acoustic frame to generate a textual token.

Career Highlights

Xu’s listed patent reflects work at the intersection of speech processing and conversational AI. By incorporating pronunciation information into token embeddings, the disclosed approach aims to help systems interpret spoken language in context.

Collaborations

Xu’s listed coworker is Zhehuai Chen. The patent information provided identifies Nvidia Corporation as Xu’s company.

Conclusion

Hainan Xu’s patent focuses on using pronunciation-aware embeddings to support the generation of text from acoustic input, highlighting a practical innovation for conversational AI.

Profile summary based on public USPTO records.
Data Sources: USPTO Patent Grant XML, Patent Center, EPO & CIPO • Normalized by IDiyas Innovation Graph. Methodology & provenance architecturePlease report any incorrect information to [email protected]
Loading…