Selected work

Representative Papers

Multi-Expert Routing for Multi-Domain Low-Resource OCR: A Manchu Case Study

Jul 15, 2026

This work addresses the challenges of diverse historical Manchu handwriting styles—such as regular script, running script, and semi-cursive memorial script—and the scarcity of annotated data in optical character recognition (OCR). The authors propose a mixture-of-experts routing system that innovatively repurposes model checkpoints from iterative fine-tuning as domain-specific experts. A lightweight page-level visual style classifier enables highly accurate expert selection, achieving 99.3% routing accuracy, and dynamically instantiates new experts when no suitable one exists. Remarkably, without access to ground-truth style labels, the system attains character error rates of 0.30%, 1.57%, and 4.83% on three test sets, matching the performance upper bound achievable with oracle style labels and substantially improving cross-style Manchu text recognition under low-resource conditions.

0 citationsRead paper

CombinationTS: A Modular Framework for Understanding Time-Series Forecasting Models

May 02, 2026

This work addresses the often statistically unstable performance gains and unclear attribution in existing time series forecasting models. It proposes CombinationTS, a framework that decouples architectures into five orthogonal modules—input transformation, embedding, encoder, decoder, and output transformation—and quantifies each component’s contribution to both predictive performance (μ) and stability (σ) under a unified evaluation protocol. Through large-scale paired experiments and probabilistic assessment, the study uncovers the “identity paradox”: with effective embeddings, a parameter-free identity encoder can match or even surpass sophisticated backbone encoders. Furthermore, it demonstrates that input transformations incorporating structural priors yield greater benefits than merely increasing encoder complexity. This work establishes a principled baseline for architectural necessity, shifting model evaluation from holistic selection toward fine-grained, component-level attribution.

0 citationsRead paper

A Similarity Network for Correlating Musical Structure to Military Strategy

Jan 18, 2026

This study addresses a critical gap in existing research by proposing an interdisciplinary analytical framework to effectively evaluate musical perception from the perspectives of system dynamics and information management, particularly in linking musical structure with complex strategic systems. Inspired by the conceptual parallels between conductors’ score interpretation and military sand-table exercises, the work introduces a novel network-based analogy between musical structures and classical military strategies such as those in Sun Tzu’s *The Art of War*. By extracting Mel-frequency cepstral coefficients (MFCCs) from film scores in war movies, the authors construct a Music Clip Correlation Network (MCCN) and apply network analysis techniques to uncover deep structural commonalities in coordination mechanisms and information organization between music and military strategy. This approach offers fresh methodological insights for both music aesthetics education and strategic thinking research.

0 citationsRead paper

Recent publications

Latest Papers

Multi-Expert Routing for Multi-Domain Low-Resource OCR: A Manchu Case Study

Jul 15, 2026

This work addresses the challenges of diverse historical Manchu handwriting styles—such as regular script, running script, and semi-cursive memorial script—and the scarcity of annotated data in optical character recognition (OCR). The authors propose a mixture-of-experts routing system that innovatively repurposes model checkpoints from iterative fine-tuning as domain-specific experts. A lightweight page-level visual style classifier enables highly accurate expert selection, achieving 99.3% routing accuracy, and dynamically instantiates new experts when no suitable one exists. Remarkably, without access to ground-truth style labels, the system attains character error rates of 0.30%, 1.57%, and 4.83% on three test sets, matching the performance upper bound achievable with oracle style labels and substantially improving cross-style Manchu text recognition under low-resource conditions.

0 citationsRead paper

CombinationTS: A Modular Framework for Understanding Time-Series Forecasting Models

May 02, 2026

This work addresses the often statistically unstable performance gains and unclear attribution in existing time series forecasting models. It proposes CombinationTS, a framework that decouples architectures into five orthogonal modules—input transformation, embedding, encoder, decoder, and output transformation—and quantifies each component’s contribution to both predictive performance (μ) and stability (σ) under a unified evaluation protocol. Through large-scale paired experiments and probabilistic assessment, the study uncovers the “identity paradox”: with effective embeddings, a parameter-free identity encoder can match or even surpass sophisticated backbone encoders. Furthermore, it demonstrates that input transformations incorporating structural priors yield greater benefits than merely increasing encoder complexity. This work establishes a principled baseline for architectural necessity, shifting model evaluation from holistic selection toward fine-grained, component-level attribution.

0 citationsRead paper

A Similarity Network for Correlating Musical Structure to Military Strategy

Jan 18, 2026

This study addresses a critical gap in existing research by proposing an interdisciplinary analytical framework to effectively evaluate musical perception from the perspectives of system dynamics and information management, particularly in linking musical structure with complex strategic systems. Inspired by the conceptual parallels between conductors’ score interpretation and military sand-table exercises, the work introduces a novel network-based analogy between musical structures and classical military strategies such as those in Sun Tzu’s *The Art of War*. By extracting Mel-frequency cepstral coefficients (MFCCs) from film scores in war movies, the authors construct a Music Clip Correlation Network (MCCN) and apply network analysis techniques to uncover deep structural commonalities in coordination mechanisms and information organization between music and military strategy. This approach offers fresh methodological insights for both music aesthetics education and strategic thinking research.

0 citationsRead paper