Finite Dependence and Invariance Hierarchies for Finitely Supported Structures
研究通过有限支持集X^S和不变性层次B^{(n)}_{G,S}探讨了有限支持结构的依赖性和不变性问题，提出并验证了几种名义原则的有效性。
0 citationsRead paper
Selected work
Representative Papers
MBO Scheme for Local Chan--Vese Segmentation
This work addresses the limited robustness of the local Chan–Vese model to intensity inhomogeneity and the computational inefficiency of conventional finite difference schemes by introducing, for the first time, the Merriman–Bence–Osher (MBO) scheme into this framework. The proposed method formulates an efficient variational level set approach grounded in local image statistics, significantly accelerating computation while naturally accommodating two-phase, multi-phase, and color image segmentation. Extensive experiments on diverse datasets—including medical and microscopic images—demonstrate that the algorithm achieves superior segmentation accuracy and speed compared to traditional finite difference methods, all while maintaining high robustness to intensity variations.
0 citationsRead paper
The GEST-Engine: From Event Graphs to Synthetic Video. A Full Technical Report
This work proposes an end-to-end framework for generating semantically coherent, spatiotemporally consistent, and multi-agent coordinated synthetic videos from natural language descriptions. The core innovation lies in the integration of a structured, executable event graph (GEST) with an explicit world model, orchestrated by a large language model (LLM) as a director agent, a procedural state backend, and a temporal constraint solver based on Allen’s interval algebra and the Floyd–Warshall algorithm. This system deterministically executes multi-agent interaction scripts within a commercial game engine, producing in a single simulation synchronized outputs including RGB video, depth maps, instance segmentation masks, skeletal poses, bounding boxes, spatial relation graphs, event-to-frame alignments, and corresponding linguistic descriptions. The approach enables zero-marginal-cost generation of densely annotated multimodal data, offering high-quality training and evaluation resources for video understanding and generation tasks.
0 citationsRead paper
Generative Augmentation of Raman Spectra for Glioma Classification
This study addresses the limitations in Raman spectroscopy–based glioma classification caused by small sample sizes, high heterogeneity, and class imbalance. To overcome these challenges, the authors propose a deep generative data augmentation approach based on β-conditional variational autoencoders (β-CVAE) to synthesize class-conditional spectral data, which is then combined with real samples for classifier training. Under rigorous patient-wise cross-validation, this strategy significantly improves classification performance for both IDH mutation status and methylation subtypes, demonstrating the efficacy of generative augmentation. Furthermore, the work explores a classification-by-reconstruction (CbR) mechanism leveraging reconstruction errors, which enhances model robustness in extremely data-scarce scenarios.
0 citationsRead paper
Machine Learning for Biomedical Raman Spectroscopy: From Spectral Acquisition to Clinical Translation
This work addresses the challenges posed by high dimensionality, strong noise, fluorescence background interference, and sample heterogeneity in biomedical Raman spectroscopy by developing an end-to-end machine learning framework that integrates preprocessing, unsupervised structure discovery, supervised diagnosis, interpretability analysis, and multimodal fusion. Innovatively combining signal correction, representation learning, transfer learning, and explainable AI techniques with pathological and molecular profiling, the framework not only enhances performance in cancer diagnosis and molecular subtyping but also prioritizes biological interpretability and clinical usability. The study further introduces a standardized pipeline, robust validation protocols, and a deployment-ready architecture, offering methodological solutions to overcome limitations arising from data scarcity, instrumental variability, and translational validation bottlenecks, thereby advancing Raman–AI systems toward reliable clinical application.
0 citationsRead paper
Recent publications
Latest Papers
Finite Dependence and Invariance Hierarchies for Finitely Supported Structures
研究通过有限支持集X^S和不变性层次B^{(n)}_{G,S}探讨了有限支持结构的依赖性和不变性问题，提出并验证了几种名义原则的有效性。
0 citationsRead paper
MBO Scheme for Local Chan--Vese Segmentation
This work addresses the limited robustness of the local Chan–Vese model to intensity inhomogeneity and the computational inefficiency of conventional finite difference schemes by introducing, for the first time, the Merriman–Bence–Osher (MBO) scheme into this framework. The proposed method formulates an efficient variational level set approach grounded in local image statistics, significantly accelerating computation while naturally accommodating two-phase, multi-phase, and color image segmentation. Extensive experiments on diverse datasets—including medical and microscopic images—demonstrate that the algorithm achieves superior segmentation accuracy and speed compared to traditional finite difference methods, all while maintaining high robustness to intensity variations.
0 citationsRead paper
The GEST-Engine: From Event Graphs to Synthetic Video. A Full Technical Report
This work proposes an end-to-end framework for generating semantically coherent, spatiotemporally consistent, and multi-agent coordinated synthetic videos from natural language descriptions. The core innovation lies in the integration of a structured, executable event graph (GEST) with an explicit world model, orchestrated by a large language model (LLM) as a director agent, a procedural state backend, and a temporal constraint solver based on Allen’s interval algebra and the Floyd–Warshall algorithm. This system deterministically executes multi-agent interaction scripts within a commercial game engine, producing in a single simulation synchronized outputs including RGB video, depth maps, instance segmentation masks, skeletal poses, bounding boxes, spatial relation graphs, event-to-frame alignments, and corresponding linguistic descriptions. The approach enables zero-marginal-cost generation of densely annotated multimodal data, offering high-quality training and evaluation resources for video understanding and generation tasks.
0 citationsRead paper
Generative Augmentation of Raman Spectra for Glioma Classification
This study addresses the limitations in Raman spectroscopy–based glioma classification caused by small sample sizes, high heterogeneity, and class imbalance. To overcome these challenges, the authors propose a deep generative data augmentation approach based on β-conditional variational autoencoders (β-CVAE) to synthesize class-conditional spectral data, which is then combined with real samples for classifier training. Under rigorous patient-wise cross-validation, this strategy significantly improves classification performance for both IDH mutation status and methylation subtypes, demonstrating the efficacy of generative augmentation. Furthermore, the work explores a classification-by-reconstruction (CbR) mechanism leveraging reconstruction errors, which enhances model robustness in extremely data-scarce scenarios.
0 citationsRead paper
Machine Learning for Biomedical Raman Spectroscopy: From Spectral Acquisition to Clinical Translation
This work addresses the challenges posed by high dimensionality, strong noise, fluorescence background interference, and sample heterogeneity in biomedical Raman spectroscopy by developing an end-to-end machine learning framework that integrates preprocessing, unsupervised structure discovery, supervised diagnosis, interpretability analysis, and multimodal fusion. Innovatively combining signal correction, representation learning, transfer learning, and explainable AI techniques with pathological and molecular profiling, the framework not only enhances performance in cancer diagnosis and molecular subtyping but also prioritizes biological interpretability and clinical usability. The study further introduces a standardized pipeline, robust validation protocols, and a deployment-ready architecture, offering methodological solutions to overcome limitations arising from data scarcity, instrumental variability, and translational validation bottlenecks, thereby advancing Raman–AI systems toward reliable clinical application.
0 citationsRead paper