Selected work

Representative Papers

An Integrated Deep Learning and Statistical Framework for Whole-Network Gene--Environment Association with Leaf Vascular Architecture

Jul 23, 2026

Traditional gene–environment association studies rely on low-dimensional vein traits, overlooking the rich structural information embedded in raw leaf images. This work addresses this limitation by treating the complete leaf venation network as a high-dimensional image-based phenotype. To enable robust analysis, the authors construct high-quality annotated data by integrating EDTER and DiffusionEdge, and propose a joint modeling framework that combines semi-parametric sparse canonical correlation analysis (SSCCA) with a truncated latent Gaussian copula to handle sparse, zero-inflated edge maps. Applied to both simulated and real-world poplar datasets, the method successfully identifies three significant gene–geography interactions, demonstrating its effectiveness and generalizability in association studies involving complex image-derived phenotypes.

0 citationsRead paper

Space-CIM: Enabling Compute-In-Memory Accelerators for Thermally-Constrained Space Platforms

Jun 04, 2026

This work addresses the thermal constraints of in-orbit data centers, where limited radiative heat dissipation restricts the performance of conventional GPUs due to their high heat density and resulting hotspots that necessitate frequency throttling. To overcome this challenge, the authors propose a “radiator-in-the-loop” co-design framework that, for the first time, jointly optimizes radiative cooling capacity with computational architecture energy efficiency. Through thermal simulations and multi-workload evaluations, they demonstrate that compute-in-memory (CIM) architectures exhibit significantly more uniform thermal distribution and higher TOPS/W efficiency compared to GPUs. Experimental results under realistic orbital thermal constraints show that CIM consistently outperforms GPUs across varying thermal budgets, thereby validating its feasibility and superiority as an AI accelerator for space applications.

0 citationsRead paper

How a Bit Becomes a Story: Semantic Steering via Differentiable Fault Injection

Dec 08, 2025

This work investigates how hardware-level single-bit flips can be leveraged to *semantically steer* vision-language model (VLM) image captioning outputs—while preserving grammatical correctness. We propose the first *semantic-differentiable fault injection* paradigm, treating bit perturbations as trainable, semantics-aware operators—moving beyond conventional fault analysis focused solely on accuracy degradation. Based on gradient sensitivity estimation, we introduce BLADE, a differentiable fault analysis framework that jointly optimizes semantic consistency and linguistic fluency of generated captions, enabling bit-level differentiation over VLM weights. Experiments demonstrate that injecting a single-bit flip suffices for fine-grained semantic steering (e.g., “dog” → “wolf”) and systematic narrative alteration. Our findings uncover the encoding structure and plasticity of semantic information at the weight-bit level in large language models, establishing a novel paradigm for robustness evaluation, adversarial defense, and interpretable AI.

0 citationsRead paper

Integrating Spatial and Temporal Effects in Seat-Belt Compliance Assessment with Telematics Data

Nov 24, 2025

Traditional roadside surveys for seat belt usage rate estimation suffer from data sparsity, temporal discontinuity, high operational costs, and inability to capture dynamic behavioral patterns or localized heterogeneity. To address these limitations, this study proposes a county-level seat belt usage rate modeling framework leveraging high-resolution telematics data. We develop a novel Bayesian multilevel regression model that jointly incorporates spatial and temporal random effects, while integrating key socioeconomic covariates—including vehicle miles traveled (VMT) and per capita income—to simultaneously account for geographic clustering and temporal dynamics. The proposed model substantially improves goodness-of-fit and inferential precision; VMT and per capita income emerge as statistically significant predictors. By overcoming the spatial coverage constraints and temporal gaps inherent in conventional surveys, our approach enables fine-grained traffic safety monitoring and provides actionable, evidence-based support for regionally targeted interventions.

0 citationsRead paper

From Service-Oriented Computing to Metaverse Services: A Framework for Inclusive and Immersive Learning for Neurodivergent Students

Sep 18, 2025

This study addresses the persistent challenges of educational accessibility and inclusivity for neurodiverse learners. We propose a metaverse-based service framework tailored to educational contexts, integrating AI-driven personalization, multimodal natural interaction, and privacy-by-design principles to realize scalable, secure, and controllable immersive virtual learning environments. Departing from conventional service computing paradigms, our approach uniquely couples adaptive learning mechanisms with metaverse infrastructure—enabling real-time, cognition-aware generation of personalized learning spaces. Empirical evaluation demonstrates statistically significant improvements in learner engagement and socio-emotional development, alongside measurable reduction in educational disparities. Furthermore, we derive actionable policy recommendations and technical implementation pathways grounded in empirical findings. This work contributes both theoretical foundations and an empirically validated architectural paradigm for building inclusive, secure, and scalable intelligent education ecosystems.

0 citationsRead paper

Recent publications

Latest Papers

An Integrated Deep Learning and Statistical Framework for Whole-Network Gene--Environment Association with Leaf Vascular Architecture

Jul 23, 2026

Traditional gene–environment association studies rely on low-dimensional vein traits, overlooking the rich structural information embedded in raw leaf images. This work addresses this limitation by treating the complete leaf venation network as a high-dimensional image-based phenotype. To enable robust analysis, the authors construct high-quality annotated data by integrating EDTER and DiffusionEdge, and propose a joint modeling framework that combines semi-parametric sparse canonical correlation analysis (SSCCA) with a truncated latent Gaussian copula to handle sparse, zero-inflated edge maps. Applied to both simulated and real-world poplar datasets, the method successfully identifies three significant gene–geography interactions, demonstrating its effectiveness and generalizability in association studies involving complex image-derived phenotypes.

0 citationsRead paper

Space-CIM: Enabling Compute-In-Memory Accelerators for Thermally-Constrained Space Platforms

Jun 04, 2026

This work addresses the thermal constraints of in-orbit data centers, where limited radiative heat dissipation restricts the performance of conventional GPUs due to their high heat density and resulting hotspots that necessitate frequency throttling. To overcome this challenge, the authors propose a “radiator-in-the-loop” co-design framework that, for the first time, jointly optimizes radiative cooling capacity with computational architecture energy efficiency. Through thermal simulations and multi-workload evaluations, they demonstrate that compute-in-memory (CIM) architectures exhibit significantly more uniform thermal distribution and higher TOPS/W efficiency compared to GPUs. Experimental results under realistic orbital thermal constraints show that CIM consistently outperforms GPUs across varying thermal budgets, thereby validating its feasibility and superiority as an AI accelerator for space applications.

0 citationsRead paper

How a Bit Becomes a Story: Semantic Steering via Differentiable Fault Injection

Dec 08, 2025

This work investigates how hardware-level single-bit flips can be leveraged to *semantically steer* vision-language model (VLM) image captioning outputs—while preserving grammatical correctness. We propose the first *semantic-differentiable fault injection* paradigm, treating bit perturbations as trainable, semantics-aware operators—moving beyond conventional fault analysis focused solely on accuracy degradation. Based on gradient sensitivity estimation, we introduce BLADE, a differentiable fault analysis framework that jointly optimizes semantic consistency and linguistic fluency of generated captions, enabling bit-level differentiation over VLM weights. Experiments demonstrate that injecting a single-bit flip suffices for fine-grained semantic steering (e.g., “dog” → “wolf”) and systematic narrative alteration. Our findings uncover the encoding structure and plasticity of semantic information at the weight-bit level in large language models, establishing a novel paradigm for robustness evaluation, adversarial defense, and interpretable AI.

0 citationsRead paper

Integrating Spatial and Temporal Effects in Seat-Belt Compliance Assessment with Telematics Data

Nov 24, 2025

Traditional roadside surveys for seat belt usage rate estimation suffer from data sparsity, temporal discontinuity, high operational costs, and inability to capture dynamic behavioral patterns or localized heterogeneity. To address these limitations, this study proposes a county-level seat belt usage rate modeling framework leveraging high-resolution telematics data. We develop a novel Bayesian multilevel regression model that jointly incorporates spatial and temporal random effects, while integrating key socioeconomic covariates—including vehicle miles traveled (VMT) and per capita income—to simultaneously account for geographic clustering and temporal dynamics. The proposed model substantially improves goodness-of-fit and inferential precision; VMT and per capita income emerge as statistically significant predictors. By overcoming the spatial coverage constraints and temporal gaps inherent in conventional surveys, our approach enables fine-grained traffic safety monitoring and provides actionable, evidence-based support for regionally targeted interventions.

0 citationsRead paper

From Service-Oriented Computing to Metaverse Services: A Framework for Inclusive and Immersive Learning for Neurodivergent Students

Sep 18, 2025

This study addresses the persistent challenges of educational accessibility and inclusivity for neurodiverse learners. We propose a metaverse-based service framework tailored to educational contexts, integrating AI-driven personalization, multimodal natural interaction, and privacy-by-design principles to realize scalable, secure, and controllable immersive virtual learning environments. Departing from conventional service computing paradigms, our approach uniquely couples adaptive learning mechanisms with metaverse infrastructure—enabling real-time, cognition-aware generation of personalized learning spaces. Empirical evaluation demonstrates statistically significant improvements in learner engagement and socio-emotional development, alongside measurable reduction in educational disparities. Furthermore, we derive actionable policy recommendations and technical implementation pathways grounded in empirical findings. This work contributes both theoretical foundations and an empirically validated architectural paradigm for building inclusive, secure, and scalable intelligent education ecosystems.

0 citationsRead paper