Xingjun Ma

Title: Innovations of Xingjun Ma in Video Recognition Technology

Introduction

Xingjun Ma is a prominent inventor based in Shanghai, China, known for his significant contributions to the field of video recognition technology. With a total of two patents to his name, he has developed innovative systems and methods that enhance the security and efficiency of video processing.

Latest Patents

Xingjun Ma's latest patents include groundbreaking work on "System and method for video backdoor attack" and "Black-box adversarial attacks on videos." The first patent outlines a comprehensive approach to generating universal adversarial trigger patterns specific to tasks, which involves a trigger generation module, an adversarial perturbation module, and a poisoning and inference module. This system optimizes trigger patterns on videos from all non-target classes while relabeling them to a target class. The second patent focuses on generating black-box adversarial attacks on video recognition models, detailing a method that includes passing input video frames through a public image model, partitioning tentative perturbations, and applying rectification weights to achieve successful attacks.

Career Highlights

Xingjun Ma is affiliated with Fudan University, where he continues to push the boundaries of research in video recognition and security. His work has garnered attention for its innovative approach to adversarial machine learning, particularly in the context of video data.

Collaborations

Xingjun collaborates with notable colleagues such as Yugang Jiang and Linxi Jiang, contributing to a dynamic research environment that fosters innovation and development in their field.

Conclusion

Xingjun Ma's contributions to video recognition technology through his patents reflect his expertise and commitment to advancing the field. His innovative methods for addressing security challenges in video processing are paving the way for future developments in this critical area.