Selected work

Representative Papers

MASCOT: Analyzing Malware Evolution Through A Well-Curated Source Code Dataset

Nov 30, 2025

Rapid malware evolution and pervasive code reuse complicate lineage inference, hindering threat attribution and defense. Method: We construct a curated dataset of 6,032 manually verified malware source-code samples and, for the first time, systematically quantify engineering attributes—including scale, development cost, code quality, security practices, and dependency structure—from a software engineering perspective. We propose a multi-view lineage analysis framework that jointly leverages code similarity, dependency graphs, and multidimensional engineering metrics to quantify association strength, reconstruct individual evolutionary trajectories, and generate interpretable lineage visualizations. Results: Empirical analysis reveals a persistent increase in malware complexity and standardization, yet significant code-quality deficiencies persist. Our approach effectively uncovers familial derivation patterns, cross-family module reuse, and ecosystem-level evolutionary dynamics—establishing a novel, engineering-informed paradigm for malware threat intelligence, attribution, and proactive defense strategy formulation.

0 citationsRead paper

A machine learning approach for image classification in synthetic aperture RADAR

Aug 06, 2025

This study addresses geometric shape and sea-ice-type classification for ground targets in synthetic aperture radar (SAR) imagery. We propose an end-to-end convolutional neural network (CNN)-based recognition framework. Methodologically, we innovatively integrate a single-scattering approximation physical model to synthesize SAR data, and systematically compare classification performance across three input modalities: simulated SAR data, reconstructed SAR images, and real Sentinel-1 acquisitions. To our knowledge, this work is the first to quantitatively analyze the impact of antenna height variation on SAR image discriminability. Experimental results demonstrate that the CNN achieves >75% accuracy on both fine-grained classification tasks, validating the effectiveness of physics-informed data synthesis in enhancing deep learning generalizability for SAR imagery. The proposed approach establishes a reproducible benchmark framework for low-sample, high-noise SAR interpretation.

0 citationsRead paper

Recent publications

Latest Papers

MASCOT: Analyzing Malware Evolution Through A Well-Curated Source Code Dataset

Nov 30, 2025

Rapid malware evolution and pervasive code reuse complicate lineage inference, hindering threat attribution and defense. Method: We construct a curated dataset of 6,032 manually verified malware source-code samples and, for the first time, systematically quantify engineering attributes—including scale, development cost, code quality, security practices, and dependency structure—from a software engineering perspective. We propose a multi-view lineage analysis framework that jointly leverages code similarity, dependency graphs, and multidimensional engineering metrics to quantify association strength, reconstruct individual evolutionary trajectories, and generate interpretable lineage visualizations. Results: Empirical analysis reveals a persistent increase in malware complexity and standardization, yet significant code-quality deficiencies persist. Our approach effectively uncovers familial derivation patterns, cross-family module reuse, and ecosystem-level evolutionary dynamics—establishing a novel, engineering-informed paradigm for malware threat intelligence, attribution, and proactive defense strategy formulation.

0 citationsRead paper

A machine learning approach for image classification in synthetic aperture RADAR

Aug 06, 2025

This study addresses geometric shape and sea-ice-type classification for ground targets in synthetic aperture radar (SAR) imagery. We propose an end-to-end convolutional neural network (CNN)-based recognition framework. Methodologically, we innovatively integrate a single-scattering approximation physical model to synthesize SAR data, and systematically compare classification performance across three input modalities: simulated SAR data, reconstructed SAR images, and real Sentinel-1 acquisitions. To our knowledge, this work is the first to quantitatively analyze the impact of antenna height variation on SAR image discriminability. Experimental results demonstrate that the CNN achieves >75% accuracy on both fine-grained classification tasks, validating the effectiveness of physics-informed data synthesis in enhancing deep learning generalizability for SAR imagery. The proposed approach establishes a reproducible benchmark framework for low-sample, high-noise SAR interpretation.

0 citationsRead paper