A General Approach to Enharmonicism
本文提出了一种基于音乐句法理论处理变音现象的一般方法，构建了由自然、半音和变音构成的抽象结构，并提出了反射和最大均匀性两个标准来评估变音系统的实用性。
0 citationsRead paper
Selected work
Representative Papers
A VLM-Enhanced Framework for Comprehensive Traffic Sign Condition Assessment Integrating Daytime Visual Performance and Nighttime Retroreflectivity Evaluation
This study addresses the limitations of conventional traffic sign assessment, which relies heavily on manual inspections and typically evaluates either daytime visibility or nighttime retroreflectivity in isolation, lacking an integrated approach. To bridge this gap, the authors propose the first unified evaluation framework that jointly quantifies daytime visual performance and nighttime retroreflective properties. Specifically, they fine-tune vision-language models—LLaVA, Qwen, and InternVL—to assess four key daytime visual factors and integrate these with LiDAR-derived retroreflectivity measurements to formulate a Sign Condition Index (SCI) for maintenance decision support. Experiments on 462 traffic signs demonstrate the framework’s effectiveness in identifying 68 signs requiring immediate replacement, with LLaVA and Qwen achieving the highest performance (bidirectional cosine similarity of 0.67–0.76), thereby validating the method’s accuracy and cost-efficiency.
0 citationsRead paper
Quantum-Enhanced Similarity Measures for Polarimetric Materials Classification
This work addresses the challenges of similarity measurement and open-set recognition in polarimetric material classification by formulating the task as a point-matching problem. The authors propose a classical-quantum hybrid framework that maps Mueller matrices to 32-dimensional voxel embeddings via an encoder, representing them as probability amplitudes of quantum states. For the first time, a quantum SWAP-test circuit is employed to compute the fidelity between query samples and anchor points, which serves as the similarity score for open-set classification. Evaluated on a dataset comprising 23 material classes, the method achieves competitive classification accuracy and demonstrates strong open-set discrimination capability.
0 citationsRead paper
Formulating Subgroup Discovery as a Quantum Optimization Problem for Network Security
This work addresses the lack of interpretability in existing network intrusion detection systems and the inability of classical subgroup discovery methods to effectively uncover critical multi-feature interactions in high-dimensional spaces. The study pioneers a quantum optimization formulation of subgroup discovery by encoding feature selection into a QUBO (Quadratic Unconstrained Binary Optimization) model, integrating least-squares regression to approximate the WRAcc (Weighted Relative Accuracy) objective function. The resulting problem is solved using the Quantum Approximate Optimization Algorithm (QAOA) on IBM Quantum hardware (ibm_pittsburgh). In experiments with 10–25 qubits, the approach achieves WRAcc scores ranging from 0.624 to 0.983 and yields unique subgroups with test accuracy up to 99.6%, substantially outperforming classical beam search. Notably, it uncovers high-order interaction patterns previously pruned by greedy strategies, establishing a novel paradigm for quantum combinatorial optimization in cybersecurity tailored to NISQ-era devices.
0 citationsRead paper
A Generalized Matrix Inverse that is Consistent with Respect to Diagonal Transformations
This work addresses the long-standing issue of inconsistency in generalized matrix inverses under nonsingular diagonal transformations—a problem particularly critical in robotics, tracking, and control systems where physical unit sensitivity is paramount. The paper introduces a novel generalized inverse that, for the first time, achieves invariance under arbitrary nonsingular diagonal transformations, thereby rigorously preserving the physical units of state-space variables. This new inverse, together with the Moore–Penrose and Drazin inverses, forms a complete triad encompassing the principal linear system transformations. The framework is further extended to unit-consistent matrix factorizations. Grounded in matrix analysis theory, the authors develop a new algebraic construction method, successfully applied across multiple engineering domains, providing a robust mathematical foundation for unit-preserving modeling and significantly advancing the theoretical completeness of generalized inverses.
0 citationsRead paper
Recent publications
Latest Papers
A General Approach to Enharmonicism
本文提出了一种基于音乐句法理论处理变音现象的一般方法，构建了由自然、半音和变音构成的抽象结构，并提出了反射和最大均匀性两个标准来评估变音系统的实用性。
0 citationsRead paper
A VLM-Enhanced Framework for Comprehensive Traffic Sign Condition Assessment Integrating Daytime Visual Performance and Nighttime Retroreflectivity Evaluation
This study addresses the limitations of conventional traffic sign assessment, which relies heavily on manual inspections and typically evaluates either daytime visibility or nighttime retroreflectivity in isolation, lacking an integrated approach. To bridge this gap, the authors propose the first unified evaluation framework that jointly quantifies daytime visual performance and nighttime retroreflective properties. Specifically, they fine-tune vision-language models—LLaVA, Qwen, and InternVL—to assess four key daytime visual factors and integrate these with LiDAR-derived retroreflectivity measurements to formulate a Sign Condition Index (SCI) for maintenance decision support. Experiments on 462 traffic signs demonstrate the framework’s effectiveness in identifying 68 signs requiring immediate replacement, with LLaVA and Qwen achieving the highest performance (bidirectional cosine similarity of 0.67–0.76), thereby validating the method’s accuracy and cost-efficiency.
0 citationsRead paper
Quantum-Enhanced Similarity Measures for Polarimetric Materials Classification
This work addresses the challenges of similarity measurement and open-set recognition in polarimetric material classification by formulating the task as a point-matching problem. The authors propose a classical-quantum hybrid framework that maps Mueller matrices to 32-dimensional voxel embeddings via an encoder, representing them as probability amplitudes of quantum states. For the first time, a quantum SWAP-test circuit is employed to compute the fidelity between query samples and anchor points, which serves as the similarity score for open-set classification. Evaluated on a dataset comprising 23 material classes, the method achieves competitive classification accuracy and demonstrates strong open-set discrimination capability.
0 citationsRead paper
Formulating Subgroup Discovery as a Quantum Optimization Problem for Network Security
This work addresses the lack of interpretability in existing network intrusion detection systems and the inability of classical subgroup discovery methods to effectively uncover critical multi-feature interactions in high-dimensional spaces. The study pioneers a quantum optimization formulation of subgroup discovery by encoding feature selection into a QUBO (Quadratic Unconstrained Binary Optimization) model, integrating least-squares regression to approximate the WRAcc (Weighted Relative Accuracy) objective function. The resulting problem is solved using the Quantum Approximate Optimization Algorithm (QAOA) on IBM Quantum hardware (ibm_pittsburgh). In experiments with 10–25 qubits, the approach achieves WRAcc scores ranging from 0.624 to 0.983 and yields unique subgroups with test accuracy up to 99.6%, substantially outperforming classical beam search. Notably, it uncovers high-order interaction patterns previously pruned by greedy strategies, establishing a novel paradigm for quantum combinatorial optimization in cybersecurity tailored to NISQ-era devices.
0 citationsRead paper
A Generalized Matrix Inverse that is Consistent with Respect to Diagonal Transformations
This work addresses the long-standing issue of inconsistency in generalized matrix inverses under nonsingular diagonal transformations—a problem particularly critical in robotics, tracking, and control systems where physical unit sensitivity is paramount. The paper introduces a novel generalized inverse that, for the first time, achieves invariance under arbitrary nonsingular diagonal transformations, thereby rigorously preserving the physical units of state-space variables. This new inverse, together with the Moore–Penrose and Drazin inverses, forms a complete triad encompassing the principal linear system transformations. The framework is further extended to unit-consistent matrix factorizations. Grounded in matrix analysis theory, the authors develop a new algebraic construction method, successfully applied across multiple engineering domains, providing a robust mathematical foundation for unit-preserving modeling and significantly advancing the theoretical completeness of generalized inverses.
0 citationsRead paper