Semi-Supervised Adaptation of Vision-Language Models for Image Classification
为解决卫星图像分类中注释样本稀缺的问题，提出SE-CLIP框架，通过半监督方式迭代挖掘高置信度样本，提高模型性能。
0 citationsRead paper
Selected work
Representative Papers
On a General Theoretical Framework for Radio Frequency Fingerprint-Based Authentication
This work addresses the lack of a general theoretical foundation for radio frequency fingerprinting (RFF) as a reliable device identifier, which has hindered its practical deployment in authentication systems. The paper proposes the first unified framework that theoretically links the physical formation mechanisms of RFFs to their core authentication properties—uniqueness, stability, distinguishability, and unforgeability. By integrating signal modeling, information-theoretic analysis, and formal security characterization, the framework systematically elucidates how RFFs are generated, evolve, and are observed across the transmitter–channel–receiver chain. This study establishes an interpretable and verifiable theoretical basis for RFF-based authentication, enabling co-design of communication and security functionalities and significantly enhancing the trustworthiness and feasibility of RFF deployment in real-world systems.
0 citationsRead paper
Recent publications
Latest Papers
Semi-Supervised Adaptation of Vision-Language Models for Image Classification
为解决卫星图像分类中注释样本稀缺的问题，提出SE-CLIP框架，通过半监督方式迭代挖掘高置信度样本，提高模型性能。
0 citationsRead paper
On a General Theoretical Framework for Radio Frequency Fingerprint-Based Authentication
This work addresses the lack of a general theoretical foundation for radio frequency fingerprinting (RFF) as a reliable device identifier, which has hindered its practical deployment in authentication systems. The paper proposes the first unified framework that theoretically links the physical formation mechanisms of RFFs to their core authentication properties—uniqueness, stability, distinguishability, and unforgeability. By integrating signal modeling, information-theoretic analysis, and formal security characterization, the framework systematically elucidates how RFFs are generated, evolve, and are observed across the transmitter–channel–receiver chain. This study establishes an interpretable and verifiable theoretical basis for RFF-based authentication, enabling co-design of communication and security functionalities and significantly enhancing the trustworthiness and feasibility of RFF deployment in real-world systems.
0 citationsRead paper