Xiaoyu Ji

Title: Innovations of Xiaoyu Ji in Image and Voice Privacy Protection

Introduction

Xiaoyu Ji is an innovative inventor based in Hangzhou, China. He has made significant contributions to the field of image and voice privacy protection. With a total of 3 patents, his work focuses on developing technologies that enhance privacy in everyday devices.

Latest Patents

Xiaoyu Ji's latest patents include a "Camera with Privacy Protection Function" and a "Method of Defending Against Inaudible Attacks on Voice Assistant Based on Machine Learning." The camera patent addresses the need for privacy protection in image acquisition. It features an image acquisition module, a parameter storage module, an access control module, and an image post-processing module. The objective is to enable images captured by the camera to resist unauthorized access, particularly from malicious third parties. This includes protection against face recognition technologies.

The second patent focuses on defending voice assistants from inaudible commands using machine learning. This method involves collecting data from both positive and negative samples, segmenting the data, and training a classifier to detect malicious voice commands. This customized approach ensures that detection and defense mechanisms are effective across different smart devices.

Career Highlights

Xiaoyu Ji is affiliated with Zhejiang University, where he continues to advance his research in privacy protection technologies. His work is crucial in addressing the growing concerns surrounding personal privacy in the digital age.

Collaborations

Xiaoyu collaborates with notable colleagues, including Wenyuan Xu and Guoming Zhang, who contribute to his research efforts and innovations.

Conclusion

Xiaoyu Ji's contributions to privacy protection through his innovative patents highlight the importance of safeguarding personal information in an increasingly connected world. His work not only enhances the functionality of devices but also addresses critical privacy concerns.