This inventor holds 1 USPTO granted patent. Top assignee: Soongsil University. Active years: 2026.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2026
Title: Doan Thien-Phuc and Deepfake Voice Detection
Introduction
Doan Thien-Phuc is an inventor associated with Soongsil University in Seoul, South Korea. Her patent work addresses the challenge of detecting artificially generated speech.
Latest Patents
The available information lists 1 patent for Doan Thien-Phuc: “Method and Apparatus for Learning Deepfake Voice Detection Model.” The disclosed method resynthesizes a real voice sample using multiple vocoders to create fake samples, forms mini-batches from real and fake samples, and applies supervised contrastive learning to a fake voice detection model to obtain a vocoder signature.
Career Highlights
Her listed invention focuses on training a model to distinguish real speech from deepfake audio by learning signatures associated with different vocoders. This approach is intended to support more effective deepfake voice detection.
Collaborations
The listed coworkers are Sou Hwan Jung and Ki Hun Hong. The patent information describes a collaborative research context connected with Soongsil University.
Conclusion
Doan Thien-Phuc’s listed patent contributes a vocoder-signature learning approach to deepfake voice detection, reflecting research associated with Soongsil University in Seoul.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile